Skip to main content

Full text of "Annals of the Missouri Botanical Garden"

See other formats




Missouri Botanical 
"" Garden 

Volume XLII 

19 5 5 

With 34 plates, 27 figures, and 60 maps 

Published quarterly at Galesburg, Illinois, by the Board of Trustees of 

the Missouri Botanical Garden, St. Louis, Mo. 

Entered as second-class matter at the post-office at Galesburg, Illinois, 

under the Act of March 3, 1879. 





of the 

Missouri Botanical Garden 

A Quarterly Journal containing Scientific Contributions from the 
Missouri Botanical Garden and the Henry Shaw School of Botany of 
Washington University in affiliation with the Missouri Botanical 



during the calendar year: February, May, September, and November. Four 
numbers constitute a volume. 

Subscription Price $10.00 per volume 

Single Numbers 2.50 each 



W. Wilson 


Lichenological Notes on the Flora of the Antarctic Continent and 
the Subantarctic Islands. I— IV 

Comparison of Juniper Populations on an Ozark Glade and Old 


Tassel Modifications in Xea Mays 

A New Species of Doryopteris from Surinam 


Selaginella rupestris and Its Allies Rolla M. Tryon 1- 99 

A New Pellaea from South Africa -Alice F. Tryon 101-102 

Studies on Asiatic Relatives of Maize _ Nalini Nirodi 103-130 

Carroll W. Dodge and Emanuel D. Rudolph 131-149 

Three New Annonaceae from Panama Robert E. Fries 151-152 

The Botanical Catalogues of Auguste de St. Hilaire John D. Dwyer 153-170 

rMarion Trufant Hall 171-194 

Norton H. Nickerson and Ernest E. Dale 195-212 

Karl U. Kramer and Rolla M. Tryon, Jr. 213-214 

A Revision of the Genus Celastrus Ding Hou 215-302 

The Food of a Hindu Village of North India (Reprinted from Bulletin 
No. 2, Bureau Statistics and Economic Research, United Provinces, 
Allahabad, 1937) Charlotte Viall Wiser, Ph.B., M.Sc. 303-412 

General Index to Volume XLII 

.4 13-418 



Emeritus Director 
George T. Moore 

Edgar Anderson 

Assistant Director 
Hugh C. Cutler 

Carroll W. Dodge, 


Robert E. Woodson, Jr., 

Senior Taxonomist 

Henry N. Andrews, 


Rolla M. Tryon, 

Assistant Curator of the 

Hugh C. Cutler, 

Curator Museum of 
Economic Plants 

George B. Van Schaack, 

Acting Curator of Herbarium 

Julian A. Steyermark, 

Honorary Research Associate 

Frederick G. Meyer, 


Alice F. Tryon, 

Research Associate 

John D. Dwyer, 

Research Associate 

Nell C. Horner, 

Librarian and Editor 
of Publications 


John S. Lehmann 


Daniel K. Catlin 

Second Vice-President 

Eugene Pettus 

Leicester B. Faust 

Dudley French 
Henry Hitchcock 

Richard J. Lockwood 

Henry B. Pflager 
A, Wessel Shapleigh 

Rorert Brookings Smith 


Arthur C. Lichtenberger, 

Bishop of the Diocese of 
Mi ssouri 

James F. Morrell, 

President of the Board of 
Education of St. Louis 

Stratford Lee Morton, 

President of the Academy of 
Science of St. Louis 

Ethan A. H. Shepley, 

Chancellor of Washington 

Raymond R. Tucker, 

Mayor of the City of St. Louis 

Hugh C. Cutler, Secretary 

Volume XLII 

Number 1 


of the 



Selaginella rupestrxs and Its Allies Rolla M. Tryon 1- 

A New Pellaea from South Africa Alice F. Tryon 101-102 





ice at G*l*»bw», IIILmoU, 

March I UN** 


of the 

Missouri Botanical Garden 


Quarterly Journal containing Scientific Contributions 

tanical Garden and the Henry Shaw School 



Washington University in affiliation with the Missouri Botanical 

• r^i 


Z^ft??". * "■ ¥?»™i Botanical Garden appears four times 


J.r!«rr"i j ^ksoum botanical Garden appears ion 

Subscription Pri< 
Single Numbers 

0.00 per volume 
2,50 each 

GAi? E N t Tre terf^T ^ °i f *"i ^ NALS ° F raE Mlssouw Botanical 
C^« v m ^ A*"* 1 *""! *"»«. published by the H. V. Wilson 


of the 

Missouri Botanical Garden 

Vol. 42 FEBRUARY, 1955 No. 1 



These are small plants, hardly exceeding a few inches in height, yet they are a 
striking part of the vegetation of the semi-arid and subalpine regions of western 
North America, covering barren soil and festooning rock ledges. The forty-three 
species included in this study comprise the section Tetragonostachys of subgenus 
Selaginella. Thirty- two species occur in America, eight in Africa and Madagascar, 
and the remainder in Asia. Superficially they appear similar and nondescript but 
under magnification they reveal structures of beauty in design and symmetry 
which amply reward the student who takes more than a casual interest in them. 

Selaginella rupestris and its allies occupy a unique place in xeric ecology for 
they are vascular plants adapted to being completely desiccated and reviving a few 
hours after moisture becomes available. Study of their physiological response to 
arid conditions would be of interest. In a number of ways the group offers data 
of value in geographic and evolutionary studies. S. rupestris is of particular cyto- 
logical and geographic significance because it has both sexual and apogamous races, 
the latter occupying a much larger territory than the former. In this species, as 
well as the others, details of the life-cycle, especially the means of fertilization and 
of dispersal, are poorly understood. Aside from botanical collectors and browsing 
by deer in lean seasons, they are relatively unaffected by animals or man and the 
habitats which most of them occupy are relatively little affected by fire, lumbering, 
grazing or cultivation. Studies of the group may be of comparative value in 
relation to species in which the activities of man have played a critical role. 

The principal effort in the present study has been directed toward a definition 
of the species and groups of species within a coherent framework of characters, to 
place the nomenclature on a firm foundation through reference to holotype material 
and to provide keys and illustrations to facilitate accurate identification. It is not 
to be expected that the keys and descriptions can be used by a student of the 
vascular plants without some introductory study of a few species, and indeed this 

Issued March 24, 1955. 


[Vol. 42 



is true of most genera. One can not successfully divorce the use of a monograph 
from specimens and it is not my intention that this treatment be wholly inde- 
pendent of such material. 

The Present Classification 

The taxonomy of the Selaginella rupestrh group is beset with a number of 
difficulties which must be recognized and resolved before an adequate classification 
can be developed. The first, and perhaps most important, is the relative plasticity 
of the characters, not only of the vegetative parts, but of the strobili, sporophylls 
and spores as well. This does not mean that taxonomically useful characters do 
not exist but that considerable caution must be used in choosing those of specific 

Relatively few species possess distinctive traits; rather most of them must be 
defined in terms of a particular combination of characters. There is evidence that 
there has been a great deal of parallel development in the group and that the 
potential variations of different structures have been realized many times. This 
situation has made it troublesome to find the morphological coherence of a species 
and the differentiating characters. Another difficulty is the very numerous char- 
acters that present themselves for consideration. In many groups, students prob- 
ably feel the characters are too few and wish for more but a superabundance of 
them can be equally trying. The inadequate nature of most herbarium specimens 
has been a source of confusion for nearly all of the early collections and many 
contemporary ones are fragmentary. Complete mats of some fifteen species were 
collected in the field and notes were made on local variations of habit and habitat. 
With the information obtained from these studies, it was possible to understand the 
relation of herbarium fragments to the living plants. The minute nature of the 
characters has undoubtedly resulted in many of them being overlooked or misin- 
terpreted. For example, the deciduous setae of S. eremophila were not detected 
until many years after the species was described. A relatively high magnification 
is needed to see the structures accurately. For general observation of the leaves 
and sporophylls 30 diameters is recommended, 40 diameters for the smaller details. 
Inasmuch as the species must largely be defined in terms of combinations of 
characters rather than distinctive ones, a set of specific characters for each species 
does not exist by itself but only in relation to other species. One set of characters 
may distinguish a species from its close relatives, additional ones, or another set, 
may distinguish it from less closely related species. It is obvious, therefore, that 
before a species can be worked out, the groups of species must be defined to estab- 
lish the circle within which character comparisons will be made. The following 
method was used to form a basis for grouping the species into natural units. 

Some twenty-five presumed species of which complete material was available 
early in the study were scored for about thirty characters. These included some 
from all parts of the plant. Their value was not known although only qualitative 
ones were used. Each species was compared with the others for the number of 



characters in common. Those that showed definite affinities with each other were 
grouped together. An analysis was then made for characters which would dis- 
tinguish between the groups. After a certain amount of modification and rear- 
rangement, the series and their characters were established. Such a method has a 
number of limitations that might make the results obtained inaccurate but when 
used with discretion it undoubtedly has value. The final test of the classification 
was the integration of some fifteen species not originally included. These took 
natural places within the system in regard to their geography, morphology and 

Although many characters had been used by previous authors for the definition 
of species these were at first disregarded to remove all prejudice as to their possible 
value. The numerous available characters were reduced to a manageable number 
in two ways. It was assumed that lineal measurements should be avoided, if at all 
possible; that qualitative differences should be used for the primary characters of 
species. Quantitative characters of real value could be used as expressed by ratios. 
Structures that were variable on individual stems and within individual mats would 
not be reliable for taxonomic use and many characters were rejected for this reason. 
When the number of possibly useful characters was reduced to convenient propor- 
tions, several species were chosen to be analyzed for the characters of specific value. 
These were species such as S. cinerascens, S. Bigelovii, S. oregana and S. tortipila of 
which I had ample material and which have a distinctive morphology, a compact 
and natural range and a rather defined ecology. Each was examined to determine 
the structures that were sufficiently stable to be used as specific characters. The 
species were then compared with each other in terms of this final set of characters 
and it was found that the entire set was usable to express species differences. Such 
characters were then used to define the species within each series. Most of the 
species thus derived were found to have the following attributes: a set of suffi- 
ciently stable characters so that the "population" showed morphological cohesion; 
complete morphological distinction from other species; a degree of morphological 
cohesion and distinction that is qualitatively and quantitatively comparable to 
other related species; a natural geographic distribution, allopatric or largely so from 
related species; and a definite local ecology although this might be rather broad 
within the whole range. These characteristics were used as criteria for determining 
the category in difficult and perplexing cases. The "good" species of the section, 
therefore, defined themselves, their attributes were determined and the other 
species were defined comparatively in similar terms. Subspecific categories were 
defined in relation to the species. 

The category subspecies is used in S. arenicola where the major variants have 
not completely developed the properties of species. The variety is used in three 
species but only one of these may be considered a proper example of the use of 
that category. That is in S. mutica where the two variants, although definite, have 



varieties are recognized largely for practical reasons, the material available being 


[Vol. 42 


insufficient to establish the proper category. In S. densa a most interesting but 
quite unsatisfactory taxonomic condition exists. The three variants are quite dis- 
tinct morphologically and geographically over large areas in the northern part of 
the range of the species. Here they would certainly rank as subspecies and perhaps 
even as species although the characters that separate them are not quite as well 
developed as in other species in the series. But the varieties all grow in the large 
central portion of the range where they exhibit such complete intergradation in 
characters that only one highly variable species could be recognized. I have chosen 
to use the category variety for them, placing perhaps a bit more emphasis upon 
their intergradation than upon their distinctiveness. 

The species descriptions have all been taken directly from the keys and are 
uniform only within each series. All of the characters mentioned in the species 
key are included in each description. Although this is a departure from the usual 
procedure, it is necessary in order to emphasize the set of characters important 
within an evolutionary line. A number of characters used in this treatment are 
peculiar to the genus or at least not commonly used in vascular taxonomy. For 
these I have tried to use suitable descriptive terms or phrases often without intend- 
ing precise definition. I quite agree with Mr. Weatherby's views that: "An attempt 
to find other characters which, even if seemingly incapable of altogether definite 
statement, could be more easily seen and better relied upon, seemed worthwhile." 1 
I have used many of the characters that he introduced and some new ones. 

Although it was not my original purpose to enter into the general classification 
of the genus in any way, it finally did seem necessary to give formal recognition 
to the major groups of species. The erection of the series has caused in turn some 
slight readjustments within the subgenus. These changes are presented in the 
synopsis of the subgenus. 

The Background of the Study 

L. M. Underwood was the first botanist to give serious attention to the Selag- 

pes iris 

His first paper, in 1898 2 , clearly expressed the condition 
U1 " 1C ""wncaoon at tnat time: "The variations of the species of Selaginella with 

many-ranked leaves have long been a puzzle to botanists Two clearly marked 

species from North America have been separated from the tangle already; 
there remain .... the widely varying forms that for the past forty years have 
found an unsatisfactory resting place under S. rupestris." Underwood described 
a total of nine new species in unraveling "the tangle"; I maintain seven of them. 

He was in correspondence with Hieronymus at Berlin and they exchanged 
specimens of many species in the group. Hieronymus did the Selaginellaceae for 
Engler & Prantl's 'Natiirlichen Pflanzenfamilien' 3 and prior to and following this 
work described many new species in the genus. In the S. r. 


*Jour. Am. Arb. 25:409. 1944. 

*Bull. Torr. Bot. Club 25:125-133. 1898. 

3 Nat. Pflanz. 1 4 :621-715. 1901. 



posed twenty-eight species 4 of which I recognize eleven. Hieronymus deserves 
credit for his general position that a multiplicity of species existed and for his 
attempts to work them out. His work suffered for reasons that were largely either 
beyond his control or else in harmony with the times. His material was scrappy 
and inadequate, sometimes hopelessly so. Also his concepts of the natural distribu- 
tion of species, particularly those of the United States, were not well developed. 
Finally, he relied largely upon quantitative characters of the leaves, setae and cilia, 
many of which have proved to be of little value. The considerable number of his 
species that are valid is probably due to the fact that he did have material of many 

The last worker to describe many species was Maxon 5 who published eight new 
ones from the United States. All but three of these I recognize as valid and two 
of those three are recognized as varieties. Maxon not only described most of the 
new species remaining in the United States but in his treatments of Pteridophyta 
for various floras of the western United States he supplied keys for their identifica- 
tion. The work of Underwood and Hieronymus was largely consolidated and 
systematized by Maxon and a knowledge of most of the species of the United States 
was placed on a firm foundation by him. His knowledge and judgment of the 
species were excellent but probably because his work in the genus was primarily 
floristic he separated the species on the most convenient characters and did not 
develop a set of characters useful for the whole group. Also, he relied in his 
descriptions too strongly, in my opinion, on quantitative characters. 

Prior to the studies of Underwood, Hieronymus and Maxon, the species were 
poorly understood. The most common policy was to place all material under the 
name of Selaginella rupestris; a few authors segregated varieties or forms of it; a 
very few segregated species. This early work is reviewed although it did not play 
an important role in the classification of the group. 

The species named Lycopodium rupestre by Linnaeus 6 was, as many he treated, 
previously well known. Among several earlier figures of the species, that of 
Dillenius 7 is the best. It presents an excellent illustration of the plant from 
Virginia. Linnaeus combined Virginian and Siberian plants in the same species 
and this broad range was enlarged upon by later authors until finally S. rupestris 
was credited with essentially a world-wide distribution. At that time it included 
many diverse species. Beauvois 8 segregated Selaginella and other genera from 
Lycopodium in 1805. He placed Lycopodium rupestre in his genus Stachy- 
gynandrum. Spring, Milde and A. Braun were the principal monographers of 
Selaginella in the nineteenth century and they recognized, to a greater or lesser 
extent, the variability within S. rupestris. Spring 9 grouped his material on the 

4 Most of them in Hedwigia 39:290-320. 1900. 

5 Most of them in Smiths. Misc. Coll. 72 s . 1920. 

6 Sp. PI. 2:1101. 1753. 

7 Hist. Muse. t. 63, fig. 11. 1741. 

8 Prodr. Aetheog. 101. 1805. 

9 Nouv. Mem. Acad. Roy. Belg. 24:57. 1850. 

[Vol. 42 


basis of habit, the rather short-stemmed northern species forming his S. rupestris 
var. borealis and the generally elongate and lax southern ones his var. tropica. 
Milde 10 used an essentially geographic segregation, recognizing as formae of S. 
rupestris material from several countries and regions. Of the ten forms he lists, I 
recognize seven as species. A. Braun 11 also recognized variants of S. rupestris as 
they occurred in Africa, using some obvious characters of the leaves, cilia and 
setae. Primarily due to floristic work, by the time of Underwood's treatment the 
following species were usually recognized as distinct from S. rupestris: S. Dregei 
Presl, S. tortipila A. Br., S. oregana D. C. Eaton and S. echinata Baker. 

The present authority on the genus as a whole, A. H. G. Alston, has treated a 
number of species in the S. rupestris group, especially as they have occurred in his 
regional treatments. Outside of the United States and Mexico the species are 
rather few and he nowhere has had to deal with a sufficiently large number to 
seriously engage the problem of specific characters. His work may be ranked with 
that of Maxon, placing our knowledge of the species of South America, Africa, 
Madagascar, India, China and northeastern Asia on a firm foundation. 

It will be noted that the species of Mexico had not been treated. This was 
probably due to the feeling that there was not sufficient material available to make 
it possible to deal successfully with the many species suspected of being present. 
The first worker to accept this difficult problem was C. A. Weatherby 12 ' 13 . In 
his two papers he was the first, from the point of view of this treatment, to derive 
a general set of specific characters and to start grouping the species into natural 
units; he was also the first to understand the special nature of the S. rupestris 
group and to adapt his classification to it. Mr. Weatherby's publications on 
Selaginella epitomize rather well his publications in systematics as a whole. His 
output was not voluminous but he had a strong predilection for the poorly under- 
stood groups and in these his abilities resulted in treatments of basic and permanent 
value. It will be apparent to students of his papers that the present study is largely 
an extension of his work. This is fitting for my interest in the group stemmed 
from his. 

Morphology and Life History 


to form a mat of more or less distinctive type. The stems of a mat may be inter- 
laced and intricate as in S. XJnderwoodii (fig. 40) and S. cinerascens or discrete 
as in S. densa (fig. 47). The branching may produce a very compact mat as in 


(tig. 49) or a loose, open one as in S. mutica. Growing conditions 
affect the form of the mat so that it is not stable in most species. Edges of cliffs, 
steep slopes, boulders and seepage crevices modify the habit of growth. In addi- 

10 Fil. Europ. Atlant. 260-263. 1867. 
11 Kuhn, Fil. Afr. 212-214. 1868. 
"Amcr. Fern Jour. 33:113-119. 1943. 
13 Jour. Am. Arb. 25:407-419. 1944. 



tion, some species may have the central stems crowded and rather erect and the 
peripheral ones closely prostrate and widely creeping. In S. tortipila the two types 
were described as separate species. A similar condition exists in S. Hansenii although 
fortunately they were not segregated. In these species the size of the mat is in- 
creased by growth of the apical buds. The older portions of the stems die pro- 
gressively toward the apical buds and strobili. The dead portions may be rather 
evenly distributed throughout the mat or, especially in the species with discrete 
branches, they may be in the center or at one side (fig. 47). A strobilus-bearing 
branch dies completely since the strobili, with rare exception, are determinate. 
Vegetative apices are theoretically indeterminate although actually they do occas- 
ionally die. 

The species of Arenicolae, with rhizomes or basal branch buds, form various 
shaped clumps or mats depending upon the extent to which the substrate permits 
or hinders the growth of the rhizomes or basal buds. S. arenicola, which usually 
grows in open sand, has quite symmetrical mats. In the species with rhizomes, as 
S. Weatherbiana, the rhizomes grow forward, occasionally branching or producing 
erect, aerial stems. These aerial branches produce a branch-system on which 
strobili are eventually produced and a short time later the whole branch-system 
dies. Most of the apices die due to the determinate strobili produced although the 
vegetative tips also die. This may be due to the death of the portion of the rhizome 
beneath the aerial branches or because of the distance the apices have grown from 
the nearest roots. New aerial branches produced by the younger parts of the 
rhizome perpetuate the plant. The leaves of the rhizome are generally similar to 
those of the aerial stems (compare figs. 7 and 8) but are usually broader, thinner, 
have a shorter seta and cilia, have a less-developed dorsal groove and the living ones 
are pale tan or a very pale green* S. arenicola and S. rupincola lack rhizomes but 
have short buds at the base of the erect aerial stems (figs. 1, 5). These buds become 
active upon the death of the branch-system; each developing into a new branch- 
system. This results in a rather bushy habit unless, as is often the case in S. 
rupincola, growth is modified by rock crevices. 

The stems often branch dichotomously but this is not the predominant type of 
branching in any of the species. The branching pattern is formed by laterals 
variously subordinate to the main stem, distance between branches and the angle 
at which they depart from the stem. Although this is obviously different in some 
species the degree to which it may be modified by growing conditions makes it an 
unreliable taxonomic character. A branch will remain subordinate to the stem 
upon which it is borne until the two apical buds are sufficiently removed from each 
other and then it will grow more rapidly. This seems to occur at about the time the 
stem and branches are separated by the death of the older portions. Branches of 
the second and third order are also subordinate to their respective main stems. In 
three species, S. utahensis, S. leucobryoides and S. asprella, the stems are fragile 
when dry and are easily broken even with careful handling. The anatomical 
nature of this character has not been investigated. 

[Vol. 42 


During periods of unfavorable growing conditions the stems become dormant. 
It is probable that many species have the ability to remain alive in the dormant 
state for many months. A specimen of S. Watsonii was planted three months 
after it had been pressed and dried and it put forth new growth. Part of a mat of 
S. densa var. dens a was planted six months after it had been collected and stored 
as an herbarium specimen and it also grew. Various changes take place when the 
stem becomes dormant. In many species the branch tips curl upward; in Eremo- 
philae they become involute. A considerable portion of the stem is involved and 
the dormant stems form ringlets in S. oregana. This curling of the stems recalls 
the behavior of S. lepidophylla and its relatives, the commercial Resurrection Plant. 
The leaves also change position, becoming more or less closely appressed to the stem. 
These changes are particularly marked in the strongly dorsiventral species in which 
the spreading lateral leaves almost completely enfold the upper ones. However, 
the erect upper leaves do not perceptibly change in position. In some species as 
S. tortipila the leaves are usually rather closely appressed in the growing state and 
change position very little when desiccated. The difference in position of the leaves 
of a species in the growing and dormant states is not sufficiently constant for gen- 
eral application. This behavior, however, has a distinct advantage to the taxonomist 
for it renders herbarium material as useful and as valid as living material for most 
characters of the plant. Conclusions as to the relation of living to herbarium 
material were confirmed by comparisons of specimens collected from species grow- 
ing in the greenhouse. There is no difference between the living dormant plant 
and the herbarium specimen prepared from it. Soaking the specimen in water will 
restore it in the same way as watering the living plant. Specimens prepared from 
actively growing plants will assume the characters of the dormant state upon 
drying with the exception that portions subject to pressure in pressing will be held 
in more or less the original position. 

I have not measured the growth of the stem of any species in its native habitat 
but measurements of material grown in the greenhouse indicate that it usually 
amounts to 1-4 cm. a year. During 100 days in the greenhouse, S. arenicola ssp. 
Kiddellii grew 30 mm. and S. viridissima grew 15 mm. although this may not 
reflect the relative growth rates of the species in their natural habitats. A speci- 

13 mm. on some of the stems. 

appears to show annual growth of 

The leafy stem is radially symmetrical in many species, that is, the leaves are 
disposed in the same manner on all sides of the stem at a given place and they are 
of the same size, shape, color and texture. This is true of the species that have 
erect stems and many of those that are prostrate. Most of the species with stems 
rather closely appressed to the ground are more or less dorsiventral. Extreme 
dorsiventral species such as S. Landii (figs. 62, 63) and S. echinata have the under 
leaves tightly appressed to the stem, the upper erect, the lateral ascending-curved, 
the under leaves longer than the upper, different in shape, thinner, tan or brownish, 
and the setae also differ. All intermediate conditions occur, some species having 



one or more dorsiventral characters so that there is not a sharp line of distinction 
between radial and dorsiventral symmetry. 

Leaves are the most useful part of the plant for purposes of taxonomy and a 
number of characters of the position, base, apex, shape, texture, cilia and seta have 
been used in this treatment. The base of the leaf, in some species, is abruptly 
adnate to the stem and distinct from it in color. In other species the leaf-base is 
decurrent, gradually blending into the stem in form and color. Red leaves often 
occur in S. Sartorii, S. Steyermarkii and S. Hansenii, especially in the area between 
the oldest living leaves and the youngest dead ones. (fig. 23, the dark portions of 
the leaves are red). These are often a bright red or tinged with purple. They 
perhaps develop color in a manner similar to the sugar maple and other deciduous 
trees in northern regions. A few double leaves have been observed. The leaves are 
firm in texture with the exception of S. oregana in which they are sufficiently soft 
and thick so that they partly collapse in drying. 

The usual sequence of leaf development appears as follows. At the beginning 
of the growing season relatively few green leaves are present at the apex of the 
stem but renewed growth of the apical bud produces new leaves and a conspicuous 
zone of green leaves develops. After a time the oldest of these die; those that were 
green at the start of the growing season die first. Under rather uniform conditions 
the death of the older leaves proceeds at a rather constant rate and probably at a 
rate similar to the production of new leaves so that a zone of green leaves of rather 
uniform length is maintained. During 100 days in the greenhouse, leaves died on 
11 mm. of stem in S. mutica and on 39 mm. in S. arenicola ssp. Riddellii. Toward 
the end of the growing season, or with the advent of a dry period, the leaves con- 
tinue to die, although no new ones are produced, until sometimes only a few green 

leaves near the bud remain. 

The apparent inhibition of growth of lateral branches has been discussed. In 
some species the leaves are affected in a similar manner. In S. cinerascens, for 
example, the leaves on the main stem are the longest and those on the primary, 
secondary and tertiary branches progressively shorter. However, once the branch 
apex is separated from the stem apex by the death of the older portions, the 
leaves that are produced are as large as those on a main stem. 

The strobilus is nearly always determinate but vegetative growth from its apex 
does occur. This is not uncommon in S. arenicola ssp. Riddellii and I have seen 
two other examples of the condition: S. X neomexicana {Slater J, US) and S. 
arizonica (A. & R. A. Nelson 1158, MO). This condition was noted in S. densa 
planted in the greenhouse in a dormant condition that produced vegetative tips oh 
two strobili upon renewal of the growth of the mat. I have seen one strobilus 
that was dichotomous (Soxman 351, US), a specimen of S. X neomexicana. In 
most species the sporophylls die progressively upward after the death of the leaves 
on the leafy branch beneath. In S. arenicola ssp. arenicola and ssp. acanthonota, 
and to some extent in S. Dregei and S. tortipila, the apical sporophylls die first and 
the basal last, and this occurs prior to the death of the leaves beneath so that the 
death of the strobilus is independent of the progressive death of the leaves. 

[Vol. 42 



The sporophylls are hinged by their stalks to the strobilus axis and when the 
spores mature and the sporangia open they move outward presumably in response 
to their water content. This movement evidently aids in the dispersal of the 
spores. Although I am not certain, the two valves of the sporangium may also 
move. I have observed in S. rupestris that megaspores will frequently remain in 
the basal sporangia after all of the others are shed, apparently because the leaves 
directly beneath do not allow the sporophylls to move sufficiently. Occasionally 
spores are found in dead strobili, and thus their liberation from the sporangium is 
not always effected. 

I have seen sporelings in only a single collection and germinated megaspores 
were never observed on the specimens. Megaspores were germinated on filter paper 
in a petri dish and then were dried. The triradial split in the spore coat and the 
gametophyte were still evident; thus if germinated megaspores had occurred in the 
collections they could have been recognized. Lyon 14 reports germination of the 
megaspores within the sporangium, and fertilization of the gametophyte prior to 
dispersal from the sporangium. I have observed nothing to support this but since 
the species do have natural ranges, many of them over considerable areas, it would 
seem that the normal life cycle is completed. Possibly germination and fertilization 
occur at rather rare intervals when the microclimate is particularly favorable. 
This may occur in the sporangium, as Miss Lyon observed, or after the spores have 
been shed. Since the stems of most species are quite tough and difficult to break, 
and are usually anchored at frequent intervals by the rhizophores and roots it does 
not seem likely that vegetative reproduction could account for the distribution of 
the species. In the group of species with fragile stems, S. utahensis, S. leucobryoides 
and S. asprella, it is not unreasonable to consider that fragments might be dispersed 
by the wind although actually they are all rather local endemics. Their restricted 
ranges, however, may be due to other causes and they may still possess an effective 
means of dispersal. 

In S. rupestris, which is apogamous throughout much of its range, fertilization 
is not a factor and the megaspores may be effectively dispersed immediately upon 
release from the sporangium. S. rupestris has the widest distribution of any species 
in recently available areas and its range may be taken to illustrate the effectiveness 
of megaspore dispersal. I have seen sporelings growing in the soil among the stems 
of a mat of S. sibirica (Calder # Billard 2Q95, MO). These are about 1 cm. long 
and inside the megaspore a foot can be seen from which are produced one to three 
roots and an erect stem. This stem does not bear leaves for the first third to half 
of its length. The first leaves are thinner, more elongate at the base and more 
widely spaced than the adult leaves. Those at the tip of one of the sporelings are 
of the adult type in aspect and disposition; the other two sporelings have only 
juvenile leaves. Two rhizophores had been produced on each sporeling, the first 
in the axil of the second or third leaf. 

14 Bot. Gaz. 32:124-141. 1901. 



The chromosome number has been determined in S. Weatherbiana (Try on & 
Tryon 5077, 2n = 18), S. mutica (Tryon & Tryon 5073, n = 9, 2n = 18) and 
in S. Bigelovii (Tryon & Tryon 5054, n = 9) . The sporophytic counts were made 
from the tips of rhizophores and the gametic ones from the contents of the mega- 
sporangia at meiosis. I am indebted to Dr. Amy Gage Skallerup for the consider- 
able work involved in making these counts. These numbers agree with the ones 
for S. selaginoides (S. spinulosa), S. helvetica and S. denticulata given by Miss 

Manton 15 as n = 9. 


The phyletic chart portrays the general relationships of the series and the species 
within them. There is not sufficient evidence of the actual course of evolution to 
more than indicate in very general terms the origin of the series. The characters 
that are of particular importance in determining a primitive or advanced phyletic 
position are the nature of the leaf-base, abruptly adnate or decurrent, the presence 
or absence of rhizomes, stolons or basal buds and the symmetry of the leafy stem. 
Arenicolae is placed as primitive because it exhibits some characters of Section 
Selaginella, the most primitive section of the subgenus. The comparison is par- 
ticularly effective between S. arenicola and S. rupincola of Arenicolae and the two 
species of Section Selaginella, S. selaginoides and S. deflexa. These four species have 
erect strobilus-bearing branches and have rhizophores borne only at the base of the 
stem. S. rupincola is not entirely uniform in the latter character in that occasional 
prostrate stems may have rhizophores borne throughout. S. deflexa is similar to 




does not mean that they are actually ancestral to the species of the series or that 
they represent the transition from Section Selaginella to Section Tetragonostachys; 
but rather that they retain the ancestral characters to a greater degree than the 
other living species. Eremophilae, on the basis of its leaf-base characters, is con- 
sidered to be the most advanced and Sartorii and Rupestres occupy a mid-position. 
Due to the limitations of space and the diversity of Sartorii it has not been possible 
to portray accurately the phyletic level of Rupestres. The radially symmetrical 
species of Sartorii are considered as more primitive than Rupestres, those with dorsi- 
ventral leafy stems and symmetrical strobili essentially on the same level and the 
species with dorsiventral strobili more advanced. 

It will be noted that the habit of growth is uniform in the Sartorii, Rupestres 



The leaf-base characters 
ve in Sartorii, Rupestres 
:s Sartorii and Rupestres 

are regarded as differentiating from Arenicolae or its ancestors, the characters of 

15 Manton, I. Problems of cytology and evolution in the Pteridophyta. p. 259. 1950. 



[Vol. 42 











26. NIVEA 





20. HANSEN] I 





















33. DENSA 

























Phyletic chart of subgenus Selaginella section Tetragonostachys. 



the leaf-base becoming fixed in each series and the same type of growth developing 
independently. The Eremophilae were derived from the Sartorii by further special- 
ization of the leaf-bases. The development of the dorsiventral leafy stem has 
occurred independently in Sartorii and Rupestres. S. Dregei, with its radially 
symmetrical leafy stem and unilateral strobilus, poses a problem. Perhaps the 
radial symmetry is primitive and S. Dregei should be placed apart from S. echinata 
and S. proxima, or perhaps, as the chart implies, it may be derived from species 
with dorsiventral leafy stems. 

The evolutionary relation of the species is usually not known although in a few 
cases it is fairly clear that one is ancestral to another, or more likely, that the two 
have had a common ancestor. Within each series the most specialized species are 
considered advanced. The species that exhibit few if any special structures and 
which may have characters in common with a less advanced series are considered 

Geography and Ecology 

Section Tetragonostachys (Map 1) occurs in temperate and tropical North 
and South America; it is absent from the Amazon basin, Central America and, with 
few exceptions, the colder regions. Its distribution is not as extensive in the Old 



Siberia. In Africa there is a single species in the region of the Atlas Mountains; 
the others occur south of the Sahara, with a better representation in eastern than 
in western Africa. Three species grow in Madagascar. The section is notably 
absent from the Malaysian-Australian region, from the Pacific Ocean and from 


Series Arenicolae (Map 2) extends more or less laterally across the southern 
United States and adjacent Mexico with an isolated species in northwestern Africa 
a distribution not unlike that of the genus Plat anus. Sartorii (Map 3) has a 
range much like that of the whole section except for its absence from boreal regions 
and from northwestern Africa. Rupestres (Map 4) is predominantly North 
American with a single species in northeastern Asia and one in the Himalayan 
region. It is the only series that has a distribution of any extent in areas of 
Pleistocene continental glaciation. The ranges of S. densa (Map 43), S. sibirica 



more glaciated territory than the others and it is undoubtedly significant that there 
it is apogamous. Eremophilae (Map 5) is New World with essentially an Andean 
distribution. It is absent in the American tropics from northern Peru to southern 


Most of the forty-three species of the section grow in the United States and 
Mexico. The two countries have a total of thirty species which represent all but 
two of the species of the New World. The United States has twenty-one species 
of which ten are endemics, while Mexico has sixteen species and five endemics. 
The southwestern United States— Texas to southern California— is the richest area 



[Vol. 42 





with a total of seventeen species. Ten species, four of them endemic, grow in 
California and eight species in Texas. There are three species in Canada, two in 
Alaska, one in Cuba, one in Guatemala and four in South America. Africa has 
five species, four of them endemics, Madagascar three endemic species, India and 
the Himalayan area three species, two endemic, and northeastern Asia two species, 
one of them endemic. 

Most of the species have continuous ranges, minor discontinuities probably 
being due to insufficient collecting. The important and apparently authentic 
examples of discontinuous range are the following. Both S. Sellowii (Map 17) 
and S. Sartorii (Map 18) are disjunct between Mexico and northern South America 
and the former species has an isolated station in Cuba. S. peruviana (Map 52) is 
similar but the disjunction is wider. S. Wightii (Map 19) grows in southern India 



Most of the species of each series are allopatric in their distribution and, with 
two exceptions, the relatively few cases of sympatric species are ecologically sep- 
arated. The examples I know in which two species of the same series actually grow 

pestris in Canada and S. peruviana 


There are two important ecological features of the species. They are mostly 
plants of dry habitats and they usually do not thrive in the presence of other plants. 
Consequently they are most abundant in regions of moderate rainfall, or in locally 
dry habitats in moist areas, and they grow in rocky, gravelly or sandy places, or on 
cliffs or barren soil where there is a minimum of other vegetation. One species, S. 
oregana, is usually an epiphyte and another, S. externa, occasionally is. Many of 
the species grow in acid soil or on acid rocks; a few of them will also grow on 


There is little 

evidence that they are important pioneers, being notably few in glaciated areas and 
in rocky places recently disturbed by fire or lumbering. Rather their chief role 
seems to be to fill the niche of locally transient but regionally permanent open, 

xeric habitats. 

The plants are not especially adapted either for the prevention of water loss or 
for its storage. The minor adaptations that are present in the form of fleshy and 
cutinized leaves are not sufficient to prevent frequent drying. The ability to 
survive desiccation undoubtedly resides in unusual physical and chemical properties 
of the cell contents. The extensive network of roots, borne close to the surface, 
enables the plants to utilize rapidly even small amounts of moisture. The species 
probably can grow in relatively moist places, as a few of them do, but because of 
their inability to compete with other plants, these habitats are not often available. 
They are restricted to the open habitats, which in a given area are usually the most 
xeric, where their special ability to survive desiccation allows them to be successful. 

[Vol. 42 


Distribution Maps 

A solid dot is used on the maps to indicate the location of a collection seen; in 
a few cases a circle is used to represent a specimen from a general region. In S. 
arenicola, half-dots are used to represent intermediates between the subspecies. 
Literature records, indicated by an X, have been used when they significantly aug- 
ment the range and when there is no doubt about the identity of the species. 
Except for S. Balansae, all of the literature records have been taken from the various 
papers of A. H. G. Alston. 


I am indebted to the curators of the herbaria of the following institutions for 
the generous loan of material. Botanisches Museum, Berlin-Dahlem, British 
Museum (Natural History), Chicago Natural History Museum, Gray Herbarium, 
Department of Botany, University of Minnesota, New York Botanical Garden, 
Museum National d'Histoire Naturelle, Paris, United States National Herbarium 
and Department of Botany of Yale University. The abbreviations for these and 
other herbaria are taken from Index Herbariorum, part I (Regnum Vegetabile, 

I wish to thank Mr. A. H. G. Alston and Mr. J. A. Crabbe of the British 
Museum, Mr. F. Ballard of Kew, Mme. Tardieu-Blot of Paris, Dr. H. Reimers of 
Berlin, and Mr. C. V. Morton of the United States National Herbarium for 
courtesies extended during the course of this study. 

Systematic Account 


Selaginella Beauv. Prod. Aethog. 101. 1805, conserved name. Type species: 
Selaginella selaginoides (L.) Link. 

Subgenus Selaginella (Section Homoeophyllae Spring, in Mart. Fl. Bras. 
1:118. 1840. Homotropae A. Br. Ind. Sem. Hort. Bot. Berol. 1857: Append. 
11. 1857, without rank. Subgenus Etiselaginella Warb. Monsunia 1:100. 1900. 
Subgenus Homoeopbylhm (Spring) Hieron. in Engl. & Prantl, Nat. Pflanz. 
1 :669. 1901.). Sporophylls uniform or similar, leaves uniform or gradually 
dimorphic. Type species: Selaginella selaginoides (L.) Link. 

Section Selaginella (Cylindrostachyae A. Br. loc. cit., without rank. 
Section Cylindrost achy s (A. Br.) Hieron. loc. cit.). Strobilus cylindrical, 
leaves spirally arranged. Type species: Selaginella selaginoides (L.) Link. 
species: L S. selaginoides (L.) Link, 2. S. deflexa Brack. 


Partly adapted from Walton & Alston, in Verdoorn, Man. Pterid. 503. 193 8. 



Section Tetrastichus (A. Br.) Tryon, stat. nov. (Tetrastichae A. Br. 

loc. cit., without rank). Strobilus tetragonous, at least the lower leaves 

decussate. Type species: Selaginella pumila (Schlecht.) Spring, species: 

1. S. pygmaea (Kaulf.) Alston (S. pumila (Schlecht.) Spring), 2. S. 

gracillima (Kze.) Alston (S. Preissiana Spring), 3. S. uliginosa (Labil.) 

Section Tetragonostachys (A.Br.) Hieron. loc. cit. (Tetragonostachyae 
A.Br. loc. cit., without rank). Strobilus tetragonous, leaves spirally ar- 



Arenicolae Tryon. Type species: Selaginella arenicola Underw. 

■pine oh 


Underw., 3. S. arenicola Underw., 3a. ssp. Riddellii (Van 
Eselt.) Tryon, 3b. ssp. arenicola, 3c. ssp. acanthonota (Underw.) 



Series Sartorii Tryon. Type species: Selaginella Sartorii Hieron. 
species: 8. S. Sellowii Hieron., 9. S. Sartorii Hieron., 10. S. Wightii 
Hieron., 10a. var. Wightii, 10b. var. Phillipsiana, 11. S. cinerascens 
A. A. Eaton, 12. S. Arsenei Weath., 13. S. macrathera Weath., 14. S. 
shakotanensis (Franch. ex Takeda) Miyabe & Kudo, 15. S. Wallacei 
Hieron., 16. S. tnutica D. C. Eaton ex Underw., 16a. var. limitanea 
Weath., 16b. var. mutica, 17. S. externa Underw., 18. S. Wrightii 
Hieron., 19. S. Steyermarkii Alston, 20. S. Hansenii Hieron., 21. S. 
carinata Tryon, 22. S. indica (Milde) Tryon, 23. S. njamnjamensis 
Hieron., 24. S. caffrorum (Milde) Hieron., 25. S. echinata Baker, 26. 
S. nivea Alston, 27. S. proxima Tryon, 28. S. Dregei (Presl) Hieron. 

Series Rupestres Tryon. Type species: Selaginella rupestris (L.) 
Spring, species: 29. S. Vardei Lev., 30. S. oregana D. C. Eaton, 31. S. 
Underwoodii Hieron., 32. S. rupestris (L.) Spring, 33. S. densa Rydb., 
33a. var. scopulorum (Maxon) Tryon, 33b. var. densa, 33c. var. 
Standleyi(Maxon) Tryon, 34. S. sibirica (Milde) Hieron., 35. S. Wat- 
sonii Underw., 36. S. utahensis Flowers, 37. S. leucobryoides Maxon, 
3 8. S. asprella Maxon. 

Series Eremophilae Tryon. Type species: Selaginella eremophila 

Maxon. species: 39. S. peruviana (Milde) Hieron., 40. S. arizonica 


Greenm. & Pf eiflf. 


[Vol. 42 



a. Stems erect or ascendent with rhizophores produced only at or near the 
base, rarely one or a few stems of a group prostrate with rhizophores pro- 
duced nearly throughout; subterranean rhizomes or ground-level stolons 
or basal branch buds present Series Arenicolae, p. 18 

a. Stems prostrate to decumbent, or irregularly ascendent, with rhizophores 
produced at or near the apex, at least of the primary stems (the branches 
may rarely be erect) , or epiphytic and long-pendent with rhizophores pro- 
duced only at the base; rhizomes and stolons absent, basal branch buds 
rarely and irregularly present, b. 

b. Leaves with the base abruptly adnate and distinct from the stem in 
color 17 on all sides of the stem, or especially on the branches, those on 
the main stem rarely decurrent Series Sartorii, p. 31 

b. Leaves with the base decurrent on all sides of the stem 17 or at least those 
on the underside strongly decurrent. c. 

c. Upper leaves with the base usually decurrent; leafy stem radially 
symmetrical to definitely dorsiventral, the zone of green leaves 18 
about equal on all sides of the stem; branch tips straight or slightly 
curled in the dormant condition Series Rupestres, p. 58 

c Upper leaves with the base abruptly adnate, distinct from the stem in 
color (rarely a distinct stem-ridge may be present); leafy stem 
strongly dorsiventral, the zone of green leaves 18 well developed on the 
upper side of the stem, very short to absent on the under side; branch 
tips involute in the dormant condition Series Eremophilae, p. 76 

Series Arenicolae Tryon, ser. nov. 

Rhizomata vel gemmae brevia simplicia ad bases caulium aeriorum hae saepe 
stolones producentes praesentia. Caules erecti vel ascendentes. Apices ramorum 
recti vel leviter curvati statu inerte. Caules frondosi radialiter symmetricales. 
bo* base decurrente vel abrupte adnata cum caule distincta colore. Typus: 


Plants terrestrial; rhizomes present, their leaves different from those of the 
aerial stem, thinner, with shorter setae, generally broader, with shorter cilia, with a 
less-developed dorsal groove, the living ones pale tan to a very pale green, or short, 
simple buds present at the base of the aerial stems, these sometimes elongated to 
form stolons or both rhizomes and basal buds present. Stems erect or ascendent, 
with rhjzophores produced only at or near the base, rarely one or a few stems 
(especially m S. ru P incola) prostrate with rhizophores produced generally through- 
° Ut; StCms Sh ° rt to m °derately so, forming distinct groups or in the species with 



tl- ^L rtM „ T : -•»—■■ "*"u urancnes DacK trom the apex of the i 

Th,s character may be obscured by death of the leaves during a Ion/ dormancy: 



rhizomes usually forming tall, loose mats. Branches short to moderately long, 
usually remote, discrete or intricate; branch tips straight or slightly curved in the 
dormant state. Leafy stems radially symmetrical, the leaves equal in position, 
length, shape and texture on all sides of the same portion of the stem, rarely 
slightly dorsi ventral in position on prostrate stems; zone of green leaves equal on 
all sides of the stem or rarely slightly longer on the upper side. Leaves with the 
base abruptly adnate and distinct from the stem in color, to decurrent; dorsal 
groove evident from base to apex, or in S. tortipila absent or poorly developed; 
setae, if present, relatively straight and stout, without a modified tip, persistent, or 
in S. tortipila with a filiform, tortuous, usually deciduous tip. 

Arenicolae is a homogeneous series although the species are quite distinct and 
show considerable diversity. The species are held together by the characters of the 
limited growth of the aerial branches and the rhizome or bud growth at the base 
of the branches. 

The leaf-base which is such an important character in the other series is variable 
here. In S. rupincola and S. Bigelovii, the only two species that show a close af- 
finity, the leaf-base is abruptly adnate to the stem and distinct in color. In S. 
Balansae, S. Weatherbiana and S. viridissima it is decurrent toward the base of the 
branches and tends to be abruptly adnate toward the tip. In S. arenicola and S. 
tortipila the leaf-bases are decurrent. The least specialized species appear to be 
S. rupincola and S. arenicola, while the most specialized, those with modified setae, 
are certainly S. tortipila and S. viridissima. No species shows an evident relation 
to a species of any other series. 


a. Leaves with the base abruptly adnate and distinct from the stem in color 
on all branches, b. 

b. Rhizomes absent; cilia of the leaves of a branch- system usually predomi- 
nantly or entirely piliform, about l / 4 as long as the width of the blade 
or longer, rarely only a few piliform; leaf -apex flat to rounded. Ari- 
zona to Texas, Tamaulipas to Sonora and southward — 1. S. rupincola, j 

b. Rhizomes present; cilia of the leaves of a branch system usually pre- 
dominantly or entirely dentiform, about l /% to % as long as the width 
of the blade, rarely some or most piliform and longer; leaf-apex usually 
carinate to broadly rounded. California and Baja California 

_„2. S. Bigelovii, p. 22 

a. All or at least the basal leaves of the aerial branches and of the rhizomes 
(when present) with the base decurrent. c. 

c. Setae of the leaves stout, relatively straight, without a modified tip, 
persistent, or setae absent; dorsal groove evident from the base to 

the apex. d. 
d. Leaves setate. e. 

[Vol. 42 


e. Rhizomes absent, short stolons occasionally present. Texas to 

Florida and North Carolina 3. S. arenicola, p. 23 

e. Rhizomes present, f. 

f. Leaf -apex flat to slightly rounded; setae tawny to whitish. 

Northwestern Africa 4. S. Balansae, p. 27 

f. Leaf -apex strongly carinate; setae lutescent to greenish-white. 

New Mexico and Colorado 5. S. Weatherbiana y p. 28 

d. Leaves muticous 6. S. viridissima, p. 28 

c. Setae of the leaves usually with a long, filiform, tortuous, often 
partly or wholly deciduous tip, rarely the tip only irregularly flexuous, 
or the setae without a modified tip; dorsal groove absent or developed 
only in the mid-portion of the blade, rarely moderately developed but 
not extending to the base or apex 7. S. tortipila, p. 29 

1. Selaginella rupincola Underw. in Bull. Torr. Bot. Club 25:129. 1898. 
(Lectotype: Woo ton 124 NY! marked by Underwood as type although he did 
not cite it as such. Para types: Palmer Q2 K; Tourney US!; Wright 21 16 NY!, 
US! cited by Underwood as 2 1 06). Figs. 1,2. Map 6. 

Selaginella Chrismarii Hieron. in Hedwigia 39:299. 1900, as Chrismari. (Holotype: von 

Chrismar B!; Purpus 3156 F, GH, MO, US is identical). 
Selaginella Chrismarii var. Karwinskyana Hieron. in Hedwigia 39:300. 1900. (Lectotype: 

Palmer 554 B! chosen because of its wide distribution although at least at GH and US 

it is mixed with S. peruviana. Paratypes: Karwinsky B!; Schaffner II B!). 
Selaginella Chrismarii var. Neeana Hieron. in Hedwigia 39:300. 1900. (Holotype: Nee 


Selaginella rupestris (L.) Spring var. rupincola (Underw.) Clute, Fern Allies, 142. 1905. 

Rhizomes absent; aerial branches with buds present at their base, erect or 
ascendent, sometimes decumbent at the base, rarely one or some of the branches of 
a plant prostrate. Leaves with the base abruptly adnate, distinct from the stem in 
color; margins ciliate, the cilia usually piliform, l / 4 as long as the width of the 
blade or longer, rarely some or most leaves with the cilia predominantly dentiform 
and shorter; apex flat to rounded to rarely carinate; setae milk-white and opaque 
or sometimes with a greenish-lutescent base, to whitish-lutescent and translucent. 

pale orange. 


pincola is closely related only to S. Bizelovii 

the characters in the key. The long spreading cilia on the leaves are most char- 
acteristic but infrequently they may be rather short. In a very few specimens that 
I have seen the stems are prostrate but these may be distinguished from species of 
other series by the basal buds. A park supervisor in Arizona informed me that in 
times of drought this plant was browsed by deer. 

Dry open places, on ledges and in crevices of cliffs, on rocky slopes or in talus, 
or in gravel, usually from 1000 to 2000 m., rarely up to 3500 m. It is apparently 
confined to igneous rocks. 





Texas to Arizona, south to Puebla. 
Representative specimens: 

United States, texas: Cory 48388 (GH); Warnock 614 (GH, US), 20893 (F> 

MO), new Mexico: Organ Mountains, Dona Ana Co., 5800 ft. July 10, 1897, Wooton 

124 (K, MO, NY, US); Wright 21 16 (sometimes as 2106) (GH, K, MO, NY, US). 

Arizona: Ferriss 26 (NY, US); Maguire 10021 (GH, US); Santa Catalina Mountains, 

April 3, 1894, Tourney (US); R. M. tf A. F. Tryon 5050 (ARIZ, B, BM, DS, F, FI, GH, 

K, MICH, MO, NY, P, RM, UC, US), 5051 (CU, GH, MIN, MO, PH, POM, US, WS). 

Mexico. Nee (B) ; 1827, Karwinsky (B). coahuila: Johnston & Muller 30J (GH). 

chihuahua: sw. Chihuahua, £. Fainter 92 in 1885 (GH, MO, US) ; £. Palmer 38 in 1908 

(B, F, GH, MO, US), sonora: Gentry 1194 (F, GH, MO, US); Wiggins tf Rollins 450 

(GH, MO, NY, US), tamaulipas: Bartlett 10529 (F, US), durango: city of Durango 

and vicinity, £. Palmer 554 in 1896 (B, BM, F, GH, MO, US); E. Palmer 485 in 1906 

(BM, F, GH, MO, US). Guanajuato: Guanajuato, 1848, von Chrismar (B). Jalisco: 

Nov. 16, 1930, M. £. Jones (BM, GH, MO, US), distrito federal: Schaffner II (B, 

GH, US), puebla: Purpus 3156 (BM, F, GH, MO, US). 

la. Selaginella X neomexicana Maxon, in Smiths. Misc. Coll. 72 5 :2. 1920, as 

o. neomexicana. 
fragment GH!. 
fragment MO!' 

S. mutica X rupincola). 
ratvoes: Wooton in 190: 


1909 US! 

1907 US! 

Map 7. 

The recognition of S. X neomexicana as a putative hybrid is based on its 
abortive spores and intermediate morphological characters. The sporangia and 

[Vol. 42 


spores are variously abortive in all of the specimens examined. No mature mega- 
spores have been seen and in only one collection (Wooton in 1904) have I seen 
microspores of a size that might indicate maturity. These are probably also abortive, 
however, since they average some 22 /x in diameter, while the microspores of S. 
rupincola and S. mutica 19 range in size from 38 to 64 fi and from 30 to 53 /i 

Although abortive sporangia and spores might be due to a number of causes, 
specimens with this character are intermediate in various degrees between the two 
proposed parental species. The erect to ascendent branches with buds at their base 
are characters of S. rupincola while the commonly linear-lanceolate leaves usually 
with a strongly carinate apex suggest S. mutica. Most specimens are rather inter- 
mediate but some, such as Wherry in 1925, are rather close to S. mutica while 
others, such as Wooton on January 9 and on September 28, 1909, are close to 
S. rupincola. 

The distribution of the hybrid is within the range of both parent species and 
it is known to grow with both at some localities. 

On open or shaded igneous rocks at about 1000-2000 m. 
Texas to Arizona. 
Representative specimens: 

United States, texas: Slater 2, 3 (US) ; Knoblock 5557 (US) ; Soxman 351 (US) ; 
April 26, 1925, Wherry (US) ; Hinckley 3404 (GH, US), new Mexico: March 3, 1907, 
Wooton (F, MO, US); Organ Mountains, Dona Ana Co., 6000 ft., Jan. 9, 1909, 
Wooton (GH, US); Sept. 28, 1902, Wooton (NY, US); Sept. 11, 1904, Wooton (US). 
Arizona: Paradise, Cochise Co., 1904, Ferriss (GH, US). 

2. Selaginella Bigelovii Underw. in Bull. Torr. Bot. Club 25:130. 1898. 
(Holotype: Bigelow NY!. Para types: Coville & Funs ton IO I NY!; Henshaw 
NY! US!; Parish & Parish 671 NY!; Underwood NY!). Fig. 3. Map 8. 

Rhizomes widely creeping; aerial branches with buds often present at their 
base, erect, rarely some decumbent at the base. Leaves with the base abruptly 
adnate, distinct from the stem in color; margins ciliate, the cilia usually dentiform, 
about y s to y 6 as long as the width of the blade, less often some or most of the 
leaves with the cilia piliform and longer; apex strongly carinate to broadly rounded, 
or predominantly so; setae usually milk-white and opaque, to whitish, lutescent- 
or greenish-white, translucent. Megaspores rugose to rugose-reticulate, rarely 
nearly smooth on the outer face, lemon-yellow to pale orange. 

lovii is closely related to S. rup 




This species usually grows in open places, on serpentine, sandstone or igneous 
rocks, in crevices or shallow depressions, on cliffs, at the base of boulders or in 

2Q Tr 7 OI1 » A - F. in Ann. Mo. Bot. Gard. 36:418, 419. 1949: 
Wiggins, I. L. in Amer. Fern Jour. 22:15. 1932. 



gravelly soil, less often in shaded or moist situations or in clay soil; it occurs from 
about sea level up to 2000 m. 

California and Baja California. 

Representative specimens: 

United States. California: Abrams 3126 (GH, MO, NY), 3402 (F, GH, MO, NY, 

US); Lt. Whipple's Exped., March 16, 1854, Bigelow (GH, NY, US, YU); Clokey 5165 

(F, GH, NY); Coville tf Funston 101 (GH, NY, US); Fosberg 697 (F, MO, NY, US); 

Heller 7276 (GH, MO, NY, US), 7610 (F, GH, MO, NY, US), 8604 (F, GH, MO, 

NY, US) ; April 3, 1893, Henshaw (NY, US) ; Moxley 876 (F, GH, NY, US) ; E. Palmer 

434 in 1875 (F, MO, NY); Parish tf Parish 671 (F, NY); R. M. tt A. F. Tryon 5053 

(BM, MO, NY, P), 5054 (F, K, MO, UC), 5056 (B, GH, MO, US); Jan. 1889, Under- 
wood (NY). 

Mexico, baja California: Ferris 8470 (NY, US) ; Wiggins 4212 (F, GH, NY, US). 

3. Selaginella arenicola Underw. in Bull. Torr. Bot. Club 25:541. 1898, 

based on S. arenaria Underw. 

Map 9. 

Rhizomes absent; aerial branches with buds present at their base, which are 
sometimes elongated to form stolons, strictly erect, ascendent to decumbent at the 
base. Leaves with the base decurrent, blending into the stem in color or sometimes 
rather abruptly decurrent on the apical portion of the stem; margins rarely eciliate, 
usually ciliate, the cilia piliform, about % as long as the width of the blade, less 
often shorter or dentiform; apex nearly flat to carinate; setae usually milk-white 
to tawny and opaque, rarely whitish with a lutescent base and translucent. Mega- 
spores rugose, rugose-reticulate to rugose-tuberculate on the commissural face, 
less prominently marked to smooth on the outer face, white to pale orange. 

The three subspecies differ in characters of essentially specific value but there 
are too many intermediates to allow them full recognition. These characters are 
those mentioned in the key to subspecies. In addition, the subspecies also tend to 
differ in various leaf characters but these are too variable within each subspecies to 
be of taxonomic value. Ssp. Kiddellii has the longest leaf-base and ssp. acanthonota 
has the shortest; ssp. acanthonota has the longest setae and ssp. Kiddellii the 
shortest; ssp. acanthonota has the stoutest leafy stem with the most leaves; ssp. 
Kiddellii has the most slender stem with the fewest leaves; and ssp. acanthonota has 
the most cilia on a leaf while ssp. Kiddellii has the fewest. Subspecies arenicola 
occupies an intermediate position in the range of variability of all of these 


Subspecies Kiddellii is evidently the most primitive of the subspecies. The 

manner of death of the strobilus is the same as that in all other species of the series 

and its spores are less distinctive. Ssp. acanthonota, with aerial rhizophores, is the 

most advanced. S. arenicola differs from S. rupincola, the other member of the 

series that lacks rhizomes, in having the base of the leaves, at least on the basal 

portions of the stems, decurrent rather than abruptly adnate. 

It grows in open habitats or in light shade, in sand or in weathered soil pockets 
of acidic rocks. It probably has considerable ability as a colonizer; at least I have 
seen it in Florida growing on the graded sides of relatively recent roads. 

North Carolina to Texas. 

[Vol. 42 



a. Aerial branches erect; all rhizophores subterranean, branched into roots 
near the base. b. 
b. Basal sporophylls of a strobilus die after the leaves beneath the strobilus; 

outer face of the megaspores usually rugose 3a. ssp. Riddellii, p. 24 

b. Basal sporophylls of a strobilus die before the leaves beneath the strobilus; 
outer face of the megaspores usually smooth 3 b. ssp. arenicola y p. 26 

a. Aerial branches erect to decumbent at the base; all or many of the rhizo- 
phores aerial and unbranched toward the base; basal sporophylls of a 
strobilus die before the leaves beneath the strobilus; outer face of the 
megaspores usually smooth 3 c. ssp. acanthonota, p. 26 

3a. Selaginella arenicola ssp. Riddel lii (Van Eselt.) Tryon, comb. nov. 

Map 10. 

Selaginella Riddellii Van Eselt. in Contrib. U.S. Nat. Herb. 20:162. 1918. (Holotype: 
Thurow 7 US!. Paratypes: Drummond 352 NY!; Holmes & Fetberolf US!; Jermy 342 
MO!, US!; Undheimer 76 MO!; Long in 1900 NY!, in 1901 NY!; Plank on Aug. 10, 



Basal buds sometimes elongated to form stolons; aerial branches erect; rhizo- 
phores subterranean, branched into roots near their base. Strobilus usually borne 
on a long leafy branch, often with vegetative apical growth; basal sporophylls 
dying after the leaves beneath the strobilus. Megaspores with the outer face usually 
rugose to rugose-reticulate, rarely smooth. 

The following material is intermediate between ssp. Riddellii and ssp. arenicola 
(Map 10): Louisiana: Dorman in 1935 (GH, US); Wherry in 1937 (US). 
Alabama: Harper 37 (GH, MO, NY, US), 3116 (MO, NY, US); Wolf in 1933 
(US); Leeds in 1934 (NY); Gattman in 1938 (F). Georgia: Harper 1800 (F, 
GH, MO, NY, US) ; Pyron & McVaugb 3051 (US) ; Duncan 8683 (MO). 

All of the material cited above from Alabama is from Eight-mile Creek and 
St. Bernard, Cullman Co., and has most of the megasporangia abortive. This is 

Co., but it may be due to other causes. 



or shallow depressions of granite, sandstone or conglomerate outcrops, or at the 
base of boulders. 

Alabama to Oklahoma and Texas. 
Representative specimens: 

United States. Alabama: Harper 3403 (GH, MO, NY, US), 3762 (GH, MO, NY, 
H?" , L ° u f A ? A: Red R iver, Hale (GH, US). Oklahoma: 1928, Draper (g Gould (US); 
Y«<rW 6426a (GH). texas: Cutler 3220 (GH, MO); Drummond 352 (GH, NY, 

XoMS" \l° 5 \ HolmeS * Fetberol f (US); Jermy 342 (MO, US); Undheimer 76 (GH, 

S^Ar^' 190 °' W - H - L ° ng (NY); Feb - 1901 > W - H - L °»S (NY); E. }. Palmer 
9540 (MO .US) 13411 (MO, US), 33364 (GH, MO); Aug. 10, 1892, Plank (NY); 

Aug. 18, 1892, Plank (NY); Reverchon 1632 (F, GH, MO, NY, US); Riddell 16 (NY); 

19 55] 



-*fc ^. 





:-; -•» 







'.* ■ 










-Jt- - 


• * 

-.4 - -\j 





[Vol. 42 


I hurow 7 (US) ; Thurow . 

(BM, F, P, MO, NY, US). 

3b. Selaginella arenicola ssp. arenicola. Fig. 4. Map 11. 

Selaginella arenaria Underw. in Bull. Torr. Bot. Club 25:129. 1898, not Baker 1 88 3. 
(Lectotype: Underwood 1355 NY! was marked by Maxon as type as to six of the 
specimens (now marked x) ; the other specimens and all of those on the sheet of the 
same collection at US are intermediate between ssp. arenicola and ssp. acanthonota. 
Paratypes: Underwood 1353a NY!; Chapman NY!. Nash 1449 is excluded as a type, 
it is ssp. acanthonota) . 

Basal buds not forming stolons; aerial branches erect; rhizophores subterranean, 

branching into roots near their base. Strobilus usually borne on a long leafy 

branch, very rarely with vegetative apical growth; basal sporophylls dying before 

the leaves beneath the strobilus. Megaspores with the outer face smooth to rarely 

Specimens that are intermediate with ssp. Riddellii have been mentioned under 
that subspecies. The following specimens are those intermediate with ssp. 
acanthonota (Map 11): Georgia: Pyron tf McVaugh 3101 (MO, US) , 31 1 4 (US) , 
3175^ (US). Florida: A. A. Eaton in 1903 (US); Underwood 1355 (NY in 

part, US). 

Subspecies arenicola grows in open places among shrubs in white sand or rarely 
in crevices or on flat exposures of sandstone or granite. 
Georgia and Florida. 
Representative specimens: 

,MO N £v t S t T q A x TES "^ E ?^^ A - H ' Curth$ 6 7'4 < GH > MO > NY > US ) I Harper 1854 
Sin i^* T h l86 ° ( >' GH ' M °' NY ' US > ; P y ron * McVaugh 3048 (US), 3 I 75 

N?' JS 5 r"^ 35I ° ( JJ°;Ii S) - FL ° RIDA: 184 °' Cb *P™* < NY > i JulyV 1934, iii« 
(NY, US); Tryonsoi4 (B, BM, GH, MO), 5 oi6 (F, MO, NY, P, US), 5 0I 7 <K, MO, 

Zs/a (NTO J3n ' 14 ' 1891 ' Vnierwooi '355 (NY, as to specimens marked x) 



Selaginella arenicola ssp. acanthonota (Underw.) Tryon, comb, nov 

Fig. 5. Map 

i: Williamson 

Selaginella acanthonota Underw. in Torreya 2:172. 1902. (Holotyp 

fragment US!. Paratype: Curtis NY!) 
Selaginella ru P estris (L.) Spring var. acanthonota (Underw.) Clute, Fern Allies, 142, 264. 

Sela ul\\lf% mh ^ If*' in Pr ° C - BioL Soc - Wash - 30:161 ' ^17. (Holotype: 
Carter rr !', fe Chapman Mtmore Herb. 3432b US!; Huger NY!; Small d 
Carter 1013 GH!, NY!; Small ef Wilson 1762 NY!; Small, Carter & Small i U Q NY!). 

Selaginella humsusa Van Esel, in Contrib. U. S. Nat. Herb.' 20:165. 1918 not Hieroi 

s.1 I! i. Hol . ot yP e: Nash 1 U9 US!. Paratypes: Eaton GH!; Rapp US!). 

Selagmella flortdana Maxon in Amer. Fern Jour. 11:1. 1921, based on S.\ 



Basal buds not forming stolons; aerial branches erect to usually decumbent at 
the base; rhizophores mostly or entirely aerial and unbranched toward their base. 

btrobilus borne on a verv nhnrt tn. rrt*A*~~+-U* 1^-^ i-.r__ i i • « 



apical growth; basal sporophylls dying before the leaves beneath the strobilus. 
Megaspores with the outer face smooth. 

Specimens intermediate with ssp. arenicola have been discussed under that sub- 
species. In the several leaf characters mentioned in the discussion of S. arenicola 
the ones characteristic of ssp. acanthonota tend to be more distinctive in the north- 
eastern portion of the range, particularly in North Carolina. 

Ssp. acanthonota grows in open places among shrubs in white sand, rarely on 


North Carolina to Florida. 
Representative specimens: 

United States, north Carolina: Blake 11481A (GH, US); Curtis (NY); Heller 
14101 (F, MO); Wilmington, July, 1892, Williamson (NY, US), south Carolina: 
Hermann IOOoS (F, GH, MO, NY, US). Georgia: Harper 1852 (F, GH, MO, NY, US), 
1957 (F, GH, MO, NY, US), 1087 (GH, MO, NY, US) ; Hermann 10128 (F, NY, US) ; 
Tryon 5027 (B, BM, GH, K, MO, P, UC, US). Florida: Chapman (Biltmore Herb. 
3432b) (GH, NY, US); Nov. & Dec. 1903, A. A. Eaton (F, GH); Foster et al. (Pi. 
Exsicc. Gray. 1305) (GH, MO, US) ; Harper 10 (F, MO, NY, US), 16 (F, GH, MO, NY, 
US), 17 (F, MO, NY, US); Jan.-Feb., 1902, Huger (NY, US); Carabelle, March 15, 
1898, Mohr (NY, US); vicinity of Eustis, Lake Co., July 16-31, 1894, Nash 1449 (F, 
GH, MO, NY, US); E. Palmer 660 in 1874 (GH, MO, US); Sept. 1902, S. Rapp (NY, 
US); Small d Carter 1013 (F, GH, NY, US); Small, Carter & Small 3349 (NY, US); 
Small (3 Wilson 1762 (NY, US) ; Tracy 7554 (F, GH, MO, NY, US) ; Tryon 5012 (BM, 
DS, GH, MIN, MO), 5021 (B, BM, F, GH, K, MO, NY, P, UC, US), 5026 (ARIZ, 
MICH, MO, RM, US), 5°93 (CU, FLAS, MO, PH, POM, US, WS). 

4. Selaginella Balansae (A. Br.) Hieron. in Hedwigia 39:318. 1900. 

Fig. 6. Map 13. 

Selaginella rupestris (L.) Spring var. Balansae A. Br. in Kuhn, Fil. Afr. 212. 1868, 
category taken from annotation of type by A. Br. (Holotype: Balansa B!. Peters, 
Mozambique, is excluded as a type; it is S. Dregei). 

Selaginella rupestris ssp. Balansae (A. Br.) f incorrectly attributed to Hieron. by] Jahand. 
& Maire, Cat. PL Maroc. 1:11. 1931. 

Rhizomes short, creeping; aerial branches with buds present at their base which 
are sometimes elongated to form stolons, erect or ascendent, rarely some decumbent. 
Leaves with the base decurrent and blending into the stem in color on the basal 
portion of the stem, rather abruptly adnate on the apical portion; margins ciliate, 
the cilia dentiform, about % or less as long as the width of the blade; apex nearly 
flat to slightly rounded; setae tawny to whitish, subopaque to translucent. Mega- 
spores rugose on the commissural face, less prominently rugose-reticulate on the 
outer face, pale yellow to pale orange. 

The unusual distribution of the series Arenicolae results from the inclusion of 
this North African species. However, the range of the series is not unlike that of 

the genus Platanus. 


cola but in technical characters it is probably closest to S. arenicola. The localities 


1:11. 1931. 

This species grows in exposed rocky places at 200-1200 m. 

[Vol. 42 

Morocco and Spanish Morocco. 
Representative specimens: 

Morocco: Gorges de l'Oued-Ghaghaia, au sud de Maroc, June 3, 1867, Balansa (B, K, 
P, NY, US); De Retz 31681 (MO); 1908, Candoger (MO); 1935, Gattefosse (US): 
1871, Hooker (NY); R. Maire 1 533 (P). 

Spanish Morocco: Quer I (BM). 

5. Selaginella Weatherbiana Tryon, in Amer. Fern Jour. 40:69. 1950. 

(Holotype: Standley 4558 MO!. Paratypes: Drouet tf Richards 3337 MO!; 

Ewan 14370 MO!; Macbride 2651 MO!; Standley 4158 US!, 4626 MO! NY! 

US!; Underwood NY!; Underwood 6 Selby 2 NY!). Figs. 7, 8. Map 14. 

Rhizomes widely creeping; aerial branches with buds rarely present at their 
base, erect or ascendent, rarely some decumbent. Leaves with the base decurrent 
and blending into the stem in color on the basal portion of the stem, less decurrent, 
sometimes rather abruptly adnate on the apical portion; margins ciliate, the cilia 
dentiform, about % as long as the width of the blade; apex carinate; setae 
lutescent- to greenish-white, translucent. Megaspores rather coarsely rugose- 


The rhizomes, the strongly carinate leaf-apex and the greenish-white or 
lutescent setae are characteristic of this species. It often grows in close association 
with S. Underwoodii and, largely on the basis of fragmentary material, was long 
confused with it. 


in crevices and on ledges of igneous rock at 2000-3200 m. 
Colorado and New Mexico. 
Representative specimens: 

United States Colorado: Ewan i 437 o (MO); Macbride 2651 (MO); R. M. & 

POM RM "uc 7 !^ 11' B < BM ' CU ' DJ \ F ' FI ' GH ' K > MICH > MIN > MO, NY, P, PH; 
POM, RM, UC US, WS); Sept. 1901, Underwood (NY); Underwood d Selby 2 (NY) 

l E Z^?°A- *"*» '7817 (US), 17980 (US); Drouet Richards 3337 $ '.MO ) 

NY Usf • 9WZ eek '/ a , TT MiSUe i ?- 800 ° ft " Wy 25 ' 1908 > Standley 4558 (F, MO 
NY, US); Standley 4158 (US), 4626 (MO, NY, US), 14692 (F, US). 

6. Selaginella viridissima Weath. in Jour. Am. Arb. 24:326. 1943. (Holo- 
type: Stewart 2204 GH!. Paratype: Johnston 8683 GH!). Fig. 9. Map 15. 

3 *t££^! A " Amer ' Fem J ° Ur - 36:5L I946 ' (Holot ^ e: C "y 44831 GH! 

Rhizomes widely creeping; aerial branches with buds rarely present at their 
base erect or ascendent, occasionally some decumbent or prostrate. Leaves with 
the base decurrent and blending into the stem in color on the basal portion of the 
stem, less decurrent, essentially distinct in color on the apical portion; margins 
ciliate cilia dentiform, % to % as long as the width of the blade; apex broadly 
rounded to carinate, muticous. Megaspores rugose, pale orange. 

The muticous leaves afford the best character to distinguish S. viridissima from 
the other species of the series. An additional character is the very slender stems. 




From additional material it is evident that S. Coryi can not be maintained on 
the basis of habit; both the Mexican and Texan material have erect or ascendent 
branches and some prostrate ones in the same mats. 

S. viridissima forms large, loose mats, rarely partly pendent, on shaded or locally 
sheltered cliffs and ledges of igneous rock. 

Southwestern Texas and Coahuila. 
Specimens examined: 

United States. Texas: Correll 13627 (MO); Chisos Mountains, Brewster Co., July 
10, 1944, Cory 44831 (GH, US) ; May 24, 1928, Murray (F) ; R. M. & A. F. Tryon 5015 
(ARIZ, B, BM, DS, F, FI, GH, K, MICH, MO, NY, P, PH, RM, UC, US) ; Warnock 
21694 (US). 

Mexico, coahuila: Johnston 8683 (GH); Kenoyer # Crum 2620 (GH) ; Canon de 
Calabasa, north wall of Sierra Maja, Oct. 27, 1941, Stewart 2204 (GH). 

>-^ l \ ■ 

• i L ■ 

■ \ \ t 

* \ I * 

* X 


« - , 


» * 


...,*- * 


\ - ^ 

f * • 

' I f ■ 


* * \ ■ 

* *m ' 


* - r 



;•• - 



m 9 i 




7. Selaginella tortipila A. Br. in Ann. Sci. Nat. V. 3:271. 1865. (Lecto- 

type: Kugel B!. Paratype: Curtis B!). 

Fig. 10. Map 16. 

Selaginella rupestris (L.) Spring var. tortipila (A. Br.) Underw. Our Nat. Ferns, ed. 3, 

140. 1888 

Selaginella Sherwoodii Underw. in Torreya 2:172. 1902. (Lectotype: Sherwood in 1902 
NY!. Paratypes: Donnell-Smith in 1882 NY!; comm. Sherwood in April, 1902 NY!; 
Sherwood in 1901 NY!. Donnell-Smith in 1881 NY! is excluded as a type, it cor- 
responds to the type of S. tortipila). 

Selaginella rupestris var. Sherwoodii (Underw.) Clute, Fern Allies, 142. 1905. 


[Vol. 42 


Rhizomes widely creeping; aerial branches with buds sometimes present at their 
base, erect or ascendent, occasionally some decumbent. Leaves with the base 
decurrent and blending into the stem in color; dorsal groove absent or developed 
only in the mid-portion of the blade; margins eciliate to ciliate, the cilia dentiform 
to piliform, up to % as long as the width of the blade; apex carinate to strongly 
carina te; setae with a filiform, tortuous to irregularly flexuous, often deciduous, 
tip, rarely lacking a modified tip, lutescent- to greenish- white, translucent to sub- 
opaque. Megaspores rugose to tuberculate, rarely smooth on the outer face, pale 
yellow to pale orange. 

The two distinctive characters of S. tortipila, the poorly developed dorsal groove 
and the tortuous, filiform seta tip, are somewhat variable. Rarely on some leaves 
of a stem the dorsal groove may be quite well developed or the setae may be nearly 
straight and stouter than usual. The strobili are usually very short, about 5 mm. 
long or less. The basal sporophylls of a strobilus usually die after the leaves 
beneath; occasionally they die before. S. arenicola is the only other species in the 
series in which the sporophylls die before the leaves beneath the strobilus, this being 



the rhizomes, a character only suggested by the many other collections available. 

Wherry 2 ^ pointed out, on the basis of field observation, that specimens de- 
scribed and identified as S. Sberwoodii were portions from the center or the upper 
part of a mat, or from mats growing in unusually exposed situations. The type 
and similar specimens of S. tortipila, on the other hand, represent material from 
the outer or lower portions of a mat or from a mat growing in a shaded place. 

S. tortipila usually forms large, compact mats on exposed granite or granite- 
gneiss cliffs or rocks; less often in thin soil in depressions of rocks or in shaded 
localities. It grows from 900-1500 m. in North Carolina to as low as 90 m. in 

North Carolina to Georgia. 

Representative specimens: 

United States, north Carolina: D. S. 6 H. B. Correll 7807 (GH, US)- 1882, 

o^oIgh m G n' ^ N J' US); 1898 ' Huger (B ' NY); ^elell 94 78 (GH, US) 

H ;/ V a \Y h °Z" m * 364 ° (MO); Broad River > 22 J" 1 ?' "41, Rugel (B, NY, US) 
Highlands, Macon Co 1902, Sherwood (NY); April, 1902, comm. Sherwood (NY) 

, , ~„^„ VJ .„ x , , Jiyiu, i^v/ii, comm. znerwt 

south Carolina: Carte (B); 1881, DonmU-Smith (F, GH, MO, NY US1 Georgia- 

$£■*"»/<&£ (GH> MO); "'""""" ,0, ° 8 <F ' GH ' MO - NY us,f,U™: 


Jour. South. Appalach. Bot. Club 1:65-69. 1936. 

alon7ttrB y ;Jl' I' Z *T ^ ^t^' *°\ Cl » b 1:68 ' ^6, says that this species grows 
along the Broad River only at Chimney Rock, Rutherford Co. Presumablv R„„1 ™/. L ^ii 

tion there. 

Presumably Rugel made his collec- 



Series Sartorii Tryon, ser. nov. 


Rhizomata et stolones nulla, gemmae breves simplices ad bases caulium rare 
praesentes vel absentes. Caules prostrati vel ramis rare ascendentibus vel erectis. 
Apices ramorum recti vel involuti statu inerte. Caules frondosi valde dorsiventrales 
vel radialiter symmetrical. Folia base abrupte adnata cum caule distincta colore 
vel folia caulium primariorum rare decurrentia. Typus: Selaginella Sartorii Hieron. 
Plants terrestrial or, in S. externa, sometimes epiphytic; rhizomes and stolons 
absent; short, simple, basal buds present only in S. Dregei, and then rarely. Stems 
prostrate with rhizophores produced generally throughout or, in S. Dregei, often 
irregularly ascendent, or in S. proxima and S. Wightii the primary branches tending 
to be erect or ascendent and dendroid with rhizophores few or absent among the 
branches; stems moderately long to long, or in S. macrathera and S. carina fa, short, 
or in S. externa, very long; usually forming loose, open, less often compact mats. 
Branches long to short, remote to approximate, usually intricate; branch tips 
straight or, in the dorsi ventral species, involute in the dormant state; the older 
primary branches of the main stem (excluding dichotomous branches) once to 
three, rarely four times pinnate. Leafy stems radially symmetrical, the leaves 
equal in length, shape, position and texture on all sides of the same portion of the 
stem, to strongly dorsiventral with the under leaves appressed, the upper erect, 
the under longer than the upper and different in shape and texture; zone of green 
leaves equal on all sides of the stem to much longer on the upper side in the dorsi- 
ventral species. Leaves with the base abruptly adnate, distinct from the stem in 
color, or those on the main stems slightly decurrent; apical portion plane or in- 
curved or, in S. externa, revolute; setae straight or nearly so, or in S. Dregei and 
S. proxima, strongly curved, entire, or in the upper leaves of S. echinata, stellate 
or substellate. Strobilus with the sporophylls in four ranks, radially symmetrical 
or dorsiventral, those on the under side longer and brcader than those on the upper 
side or, in S. Dregei, with the sporophylls in two ranks on the under side of the 
branch tip. 

This is the most heterogeneous series consisting of three species groups that in 
a more elaborate classification might be given formal recognition. The least special- 
ized and presumably most primitive group, S. Sellowii to S. externa, has the leafy 
stems and strobili radially symmetrical. The first three species, S. Sellowii, S. 
Sartorii and S. Wightii, are quite generalized in their characters, the others are 
specialized in one or more ways. The next group, S. Wrightii to S. indica, has 
dorsiventral leafy stems and radially symmetrical strobili. All of the species are 
specialized in some character. The last and most marked group, S. njamnjamensis 
to S. Dregei, has the leafy stems radially symmetrical or dorsiventral and the strobili 
dorsiventral. These are probably the most highly evolved species in the section 
and among them S. echinata of Madagascar stands as the most specialized. The 
character of the dorsiventral strobilus has been observed very rarely in S. viutica 
of this series and in S. Parishii in Eremophilae. 

[Vol. 42 



Nearly all of the species of the series are quite distinctive and many of them 
are exceptionally so. The few close relations that afford a critical separation are 
S. Selloivii and S. Sartorii, S. njamnjamensis and S. caffrorum and perhaps S. ciner- 
ascens and S. Arsenei. 

The following specimen may represent a new species of the Sartorii but the 
material is sterile and thus inadequate: M'pala ad litt. occid. lacus Tanganyika, 
1898, Rev. Guilleme 8 (P). It may be a variant of S. Wightii var. Phillipsiana to 
which it is evidently related. Its most distinctive character is the margin of the 
leaf that is smooth or bears a few minute, dentiform cilia. An annotation label by 
Hieronymus notes it as a new species bearing the name of the collector. 


a. Sporophylls in four ranks, the strobilus radially symmetrical. New 
World, Asia, 1 species in eastern Africa, b. 

b. Leafy stems radially symmetrical or nearly so, or dorsiventral only in 
the position of the leaves, the upper and under leaves equal in length 
and shape or very nearly so on the same portion of the stem. c. 

c. Apical portion of leaves essentially plane or incurved, d. 

d. Apex of the blade plane or gently beveled in profile, flat to 
rounded, e. 

e. Cilia, especially of the leaves of the main stems, piliform, or 
dentiform only toward the apex of the blade; or if all cilia 
dentiform then the setae milk-white, opaque; or leaves eciliate. f. 
f. Leaves definitely setate. g. 

g. Older primary branches once to rarely twice pinnate; apex 
of the leaves usually convexly acute to acuminate. New 
World, h. 

h. Leaf -bases glabrous; setae usually milk-white, opaque, 
if whitish to greenish-lutescent and subtranslucent 

(Brazil) then slightly attenuate. Mexico 

Sellowiu v. 34 

h. Leaf-bases predominantly pubescent, or if rarely mostly 
glabrous then the setae strongly attenuate; setae whitish 
to greenish to lutescent, translucent to subopaque. 

Sartorii. p. 36 

g. Older primary branches twice to three times pinnate; apex 

of the blade long-acuminate. India and Africa 

10. S. Wightii, p. 39 

f. Leaves, especially of the growing tips of the secondary 
branches, muticous or with a slightly modified apex. Southern 

California and Baja California 11. S. cinerascens, p. 40 

e. Cilia of the leaves dentiform; setae whitish or greenish- to 
lutescent-whitish, translucent to subopaque. Mexico, i. 



i. Setae about % or less as long as the blade 12. S. Arsenei, p. 41 

i. Setae % to usually % as long as the blade_.13. S. macrathera, p. 41 
d. Apex of the blade predominantly truncate to abruptly beveled in 

profile, especially those of the growing-tips, broadly rounded to 

carina te. j. 

j. Apex of the sporophylls carinate; leafy stem somewhat dorsi- 

ventral in the position of the leaves. Japan.. 14. S. shako tanensis, p. 43 
j. Apex of the sporophylls rounded to broadly rounded, or if 
carinate then the leafy stem quite or essentially radially sym- 
metrical, k. 
k. Leaves ligulate to ligulate-long-triangular, definitely setate. 

Montana to British Columbia and California- 15. S. Wallace?, p. 43 
k. Leaves oblong- to ovate-lanceolate, muticous or nearly so in 

Colorado and Utah, setate southward 16. S. mutica y p. 44 

c. Apical portion of many leaves revolute. Mexico 17. S. extensa, p. 47 

b. Leafy stems strongly dorsiventral, the under leaves with the blade defi- 
nitely longer than the upper on the same portion of the stem, or dif- 
ferent in shape, or in both length and shape. 1. 

1. Sporophylls long-ciliate toward the base; under leaves linear; setae 
arising very evenly in form and usually in color from the blade-apex 

or the leaves muticous 18. S. Wrightii, p. 48 

1. Sporophylls eciliate to short-ciliate toward the base. m. 

m. Setae arising rather abruptly in form and color from the blade- 
apex. New World, n. 
n. Apex of the blade flat to rounded; some red leaves usually 
present, especially among the under ones. o. 
o. Apex of the blade of the upper leaves long-acuminate; under 

leaves papyraceous. Guatemala and adjacent Mexico 

19. S. Steyermarkii, p. 49 

o. Apex of the blade of the upper leaves acute to convexly acute; 

under leaves herbaceous. California 20. S. Hansenii, p. 49 

n. Apex of the blade carinate; no red leaves present. Ecuador 

21. S. carinata, p. 50 

m. Setae arising evenly in form and usually also in color from the 

blade- apex; under leaves papyraceous. India 22. S. indica, p. 52 

a. Sporophylls in two ranks, or in four ranks and the strobilus dorsiventral, 
the sporophylls of the under side longer and broader than those of the 
upper side. Africa and Madagascar, p. 

p. Sporophylls in four ranks; longest cilia of the leaves up to l /z as long as 
the width of the blade, q. 
q. Setae whitish to tawny or lutescent, entire. Africa, r. 

[Vol. 42 


r. Under leaves herbaceous, the base abruptly adnate, distinct from 
the stem in color; setae poorly differentiated from the blade-apex 

in form and color 23. S. njamnjamensis, p. 52 

r. Under leaves papyraceous to papyraceous-herbaceous, the base 
often decurrent, blending into the stem in color on the main stem; 
setae predominantly whitish, rather abruptly distinct from the 
blade-apex in form and color 24. S. caffrorum, p. 53 

q. Setae milk-white, opaque, entire, or those of the upper leaves, espe- 
cially at the branch-tip, whitish, substellate, soon deciduous. Mada- 
gascar, s. 

s. Setae of upper leaves substellate, soon deciduous__2 5. S. echinata, p. 54 
s. Setae entire, persistent, t. 

t. Setae straight, l / 4 to % as long as the blade 26. S. nivea, p. 55 

t. Setae strongly curved, % to l /z as long as the blade 

27. S. proxima, p. 56 

p. Sporophylls in two ranks; longest cilia of the leaves l /z as long to as 
long as the width of the blade; setae usually strongly curved. Africa 
28. S. Dregei y p. 57 

8. Selaginella Sellowii Hieron. in Hedwigia 39:306. 1900, maintained over 
S. Arechavaletae and S. montevideensis by Alston, in Physis 15:252. 1939. 
(Lectotype: Sellow in 1821 B! fragment NY!; Vie 232 US is nearly identical. 
Paratypes: De Moura 908 B; Rio de Janeiro, Gaudichaud B; Kiedel 7 B; 
Schwache 949 B; Lhotzky 7 B) . Fig. 11. Map 17. 

Selaginella rupestris (L.) Spring f. amazonica Milde, Fil. Europ. Atlant. 263. 1867. Holo- 

type: Humboldt & Bonpland B!; Peru, Galathea Exped. US 619317 is an excellent 

Selaginella rupestris f. brasiliensh Milde, Fil. Europ. Atlant. 263. 1867. (Lectotype: 

Sellow in 1821 B! fragment NY!. Paratype: Sellow B! fragment NY!. Moritz 370 

B! is excluded as a type, it is S. Sartorii). 
Selaginella rupestris var. brasiliensis (Milde) Hieron. in Engl. Bot. Jahrb. 22:417. 1896, 

not S. brasiliensis (Raddi) A. Br. 1865. 
Selaginella montevideensis Hieron. in Hedwigia 39:309. 1900. (Lectotype: Sellow 6/0 

B! fragment NY!; Gibert 1323 has been annotated at NY as a close match. Paratype: 

Gaudichaud B!). 

Selaginella amazonica (Milde) Hieron. in Hedwigia 39:310. 1900, not S. amazonica 
Spring, 1840. 

Selaginella Arechavaletae Hieron. in Hedwigia 39:311. 1900. (Holotype: Arechavdeta 
472 B! fragment NY!; Herter, Pi. Urug. 2077 is a close match). 

Selaginella Mildei Hieron. in Engl. & Prantl, Nat. Pflanz. 1 4 :671. 1901, based on S. 
amazonica (Milde) Hieron. 

Selaginella pauciciliata Hieron. in Hedwigia 58:284. 1916. (Holotype: "Wright 18 20 B!, 

Wright f 20' cited by Hieron. is actually 1820 as indicated by an extra label on the 
sheet at GH). 

Stems with the older primary branches once to rarely twice pinnate; leafy stems 
radially symmetrical to rarely somewhat dorsiventral in position, length and shape 
of the upper and under leaves. Leaves herbaceous, none red, subulate-long- 



triangular to broadly ligulate to lanceolate; base abruptly adnate, distinct from 
the stem in color or sometimes rather decurrent on the leader stem, glabrous; 
margins eciliate or usually ciliate, the cilia usually piliform, dentiform only at the 




% to nearly l / 2 

long as the blade, slightly attenuate, milk-white and opaque to greenish- or whitish- 
lutescent and subopaque to translucent. Sporophylls with the margins eciliate to 
moderately long- ciliate toward the base; apex rounded. Megaspores rugose to 
rugose-reticulate on the commissural face, rugose-reticulate to smooth on the outer 
face, yellow to orange-yellow. 

This species and the next, S. Sartorii, are not well separated by distinctive 
morphological characters although in the regions where they both occur, Mexico, 
Colombia and Venezuela, the distinctions are adequate. Each is sufficiently vari- 
able, however, so that they overlap in the total range of variation of the characters. 
I would regard the two as something more than subspecies and yet less than species 
as judged by the attributes of the other species of the series. However, since they 
are clearly separable it seems best to give them specific status. S. Sellowii is char- 
acterized by glabrous leaf-bases and milk-white, opaque setae, while S. Sartorii has 
pubescent leaf-bases and whitish to greenish-lutescent and translucent to sub- 
opaque setae. Some Brazilian material of S. Sellowii is exceptional in having setae 
characters like S. Sartorii and rarely some Mexican material of S. Sartorii will have 
glabrous leaf-bases. 

S. Sellowii is composed of three major variants all of which intergrade rather 
freely. The value of the characters and particularly their degree of correlation is 
less than in S. mutica, for example, where varieties are maintained. For these 
reasons, the variants are not accorded a status. 

The most common variant has the leafy stem radially symmetrical, the leaves 
narrowly long-triangular to long-triangular with the apex plane in profile, with 
few or no cilia and with the setae milk-white and opaque. This ranges from Bahia, 
Brazil, to Argentina and Bolivia with slight variants extending northward to 

Mexico and Cuba. 


the leaves piliform-ciliate; that from Mexico and Cuba has the leafy stem more 




S. montevideensis are this variant although those of the last three are somewhat 
intermediate with the next. 

The second variant is usually slightly dorsiventral; the leaves are broadly 
ligulate to ligulate-lanceolate with the apex slightly beveled in profile, with num- 
erous cilia and with the setae whitish and often subopaque. It grows from Rio 
Grande do Sul, Brazil, to Argentina and Peru. The type specimen of S. Mildei 
belongs here. 


[Vol. 42 

The third variant is similar to the second but the leafy stem is very slender; 
the leaves are not as broad and the setae are greenish- to whitish-lutescent and 
translucent to subopaque. It is known from the state of Rio de Janeiro, Brazil, 
and from Rio La Plata (Settling, GH). 

It is evident that Hieronymus essentially based S. Sellowii on S. rupestris f. 
brasiliensis, the epithet brasiliensis having been previously used at the specific level. 
However, he did not do it precisely so I am choosing the same type specimen for 
both names. This species and S. Sartorii are both quite generalized in their char- 
acters and probably represent the basic New World type of this series. The localities 
taken from the literature (Map 17) are from Alston, in Physis 15:253. 1939. 

On exposed or wooded rocky bluffs and cliffs or among stones, less often in 
sandy or clay soil, usually at about 1000-2000 m. but ranging from sea level to 
2700 m. 

Central Mexico; Cuba; Venezuela to Colombia, Argentina and eastern Brazil. 
Representative specimens: 

Cuba: Mayari, Holquin, Aug. 17, 1859, 1860, C. Wright 1820 (B, GH, K, MO, NY, 

Mexico, distrito federal: Rose 6 Painter 6835 (US), puebla: Arsene 891 (US), 
10639 (US). Veracruz: Liebmann 2062 (US). 
Venezuela: Tamayo 298 (US). 
Colombia: Lehmann KK97 (US), BT776 (US). 
Ecuador: Jameson 426 (US); Sodiro 47/1 (B). 

Peru: Biles 608 (US); Herrera 3009a (US); ad ripam fluminis Maranon, Humboldt 
8 Bonpland (B) . 

Bolivia: Balls 6229 (US); Eyerdam 25041 (F, GH) ; Steinbach 8254 (F, GH, MO), 
8574 (F, GH, MO) ; R. S. Williams 1 400 (US) . 

Argentina: Bartlett 20456 (US); Burkart 809 (GH), 3308 (GH), 7128 (GH), 
ion 3 (MO) ; Burkart tf Troncoso 11174 (MO) ; Lossen 246 (F, MO) ; T. Meyer 2227 
(GH), 3687 (NY, US); Schulz 753 (GH); L. B. Smith 4679 (US); Venturi 1 144 (GH, 

Paraguay: Fiebrig 4907 (GH, US) ; Hassler 12535^/2 (US) ; Jorgensen 4073 (F, NY) ; 
Lindman A2319 (US). 

Uruguay: Montevideo, June, 1876, Arechavaleta 472 (B, K, NY); Montevideo, 
Gaudichaud {^);Gibert 1323 (NY, US); Herter 76113 (GH, MO), 78856 (F), 04884 
(MO), 99600 (MO); Montevideo, Sellow 670 (B, K, NY). 

Brazil: Archer tf Gehrt 118 (US); Dusin 14 (US); Glaziou 3306 (US); Jurgens 29 
(NY), 308 (US), 309 (US); Undman A213 (US); Lutzelberg 18870 (NY); Riedel 7 
(GH); Rose & Russell 19937 (NY, US); Sellow (B, NY), Praia de San Diego, 1821, 
Sellow (B, NY) ; Vie 232 (US) ; Wilkes, U. S. So. Pacific Expl. Exped. (GH, NY, US, 
YU); 1934, S. Wright (GH). 

9. Selaginella Sartorii Hieron. in Hedwigia 39:304. 1900. (Lectotype by 
Weath. in Jour. Arn. Arb. 25:418. 1944: Sartorius B!. Paratvpes: Liebmann 

in Aug. 1841 B; Schiede in May, 1829 B). 


Selaginella^ Sartorii var. oregonensis Hieron. in Hedwigia 39:305. 1900. (Holotype: 

"Lyall" B! fragment NY!, collector and locality unknown). 
Selaginella Sartorii var. venezuelensis Hieron. in Hedwigia 39:305. 1900. (Holotvne: 

Moritz 370 B\) . 




[Vol. 42 


Selaginella rupestris (L.) Spring var. Sartorii (Hieron.) Frye, Ferns Northwest, 31. 1934. 

Setaginella porrecta Weath. in Jour. Arn. Arb. 25:416. 1944. (Holotype: Pennell 17198 

US! fragment GH!. Paratypes: Bar tie it 1 03 5 5 US!, 1 1039 US!; Or cut t 1098 US!, 

1142 US!, J 
US!. Schaff 


in Jour. Arn. Arb. 25:418. 1944. (Holotype: Hint on 8423 
GH! fragment US!). 

Stems with the older primary branches once to rarely twice pinnate; leafy stems 
radially symmetrical except in the position of the leaves, or very slightly dorsi- 
ventral in the position and shape of the leaves. Leaves herbaceous to herbaceous- 
papyraceous, often some red, ligulate to ligulate-long-triangular; base abruptly 
adnate, distinct from the stem in color or sometimes slightly decurrent on the 
leader stems, usually pubescent, rarely glabrous; margins ciliate, the cilia piliform, 
or dentiform only toward the apex of the blade, the longest cilia l / 4 to l /z as long 
as the width of the blade; apex acuminate to convexly acute, flat to slightly 
rounded, plane to gently beveled in profile; setae arising rather evenly in form but 
less so in color, l / 4 (rarely less) to nearly l / 2 as long as the blade, slightly to strongly 
attenuate, whitish to greenish to lutescent, subopaque to translucent. Sporophylls 
with the margins short- to rather long-ciliate toward the base; apex slightly 
rounded to carinate. Megaspores prominently and irregularly rugose to rugose- 
reticulate on the commissural face, more prominently marked to nearly smooth 
on the outer face, yellow to orange-yellow. 

S. Sartorii is not strongly separated from S. Sellowiu The Venezuelan specimens 
of S. Sartorii, for example, are identified by the pubescent leaf-bases, a character 
that is usually variable within a species. They also differ from the S. Sellowii of 
that area, however, in the color and opaqueness of the setae. The relations of the 
two species are discussed in greater detail under S. Sellowii. 

Some of the Colombian and Venezuelan specimens such as Lindig 1523, Amor- 
tegui A6l and Tamayo 1753 have some of the leaves with a rather decurrent leaf- 
base. Hint on 3453 (BM) has the leaves almost entirely dentiform-cilia te but 
otherwise it agrees with S. Sartorii. 

Like S. Sellowii, this species is a complex but it is not as variable nor are the 

extremes as well marked. 

Hintonii and S. porrecta 

the basis of characters of specific value in other species in the series. Without 
more adequate collections it is not clear whether or not they represent geographic 




Exposed cliffs and rocky slopes of igneous rock, sandstone or limestone at 
870-2000 m. 

Mexico, Colombia, Venezuela. 
Representative specimens: 



Mexico, tamaulipas: Bartlett 10355 (US), 11039 (US); Meyer & Rogers 31 12 
(MO) ; Stanford et al. 2187 (MO), nuevo leon: Orcutt 1098 (US), 1 142 (US), 1151a 
(US); "Alamar", Pablillo, southeast of Galeana, 1650-1700 m., July 2-3, 1934, Pennell 
17198 (GH, US); Tharp 1792 (US), san luis potosi: Orcutt 5125 (US); Pennell 
J 7753 (US), durango: Ortega 5319 (US). Mexico: Ypericones, Temascaltcpcc, Sept. 
7, 1935, Hinton 8423 (GH, US). Veracruz: Mirador, Sartorius (B); 1841, Liebmann 
(GH) ; Purpus 8463 (GH, MO, US), oaxaca: Andrieux 2 (GH) ; Kenoyer 1566 (GH). 

Colombia: Amortegui A61 (US); Lindig 1523 (GH) ; Purdie (YU). 

Venezuela: Paramo de Mucuchies, Moritz 370 (B, K, YU); Tamayo 1753 (US), 
J 773 (US). 

10. Selaginella Wightii Hieron. in Hedwigia 39:319. 1900. (Lectotype: 

Wight IS B!. Para types: Walker B; Wight 283 B, JJp^ B!). Map 19. 

Stems with the primary branches tending to be ascendent and dendroid, the 
older primary branches twice to three times pinnate; leafy stems radially sym- 
metrical or nearly so. Leaves herbaceous to herbaceous-papyraceous, none red, 
long- triangular to ligulate-long-triangular; base abruptly adnate, distinct from 
the stem in color, glabrous; margins cilia te, the cilia piliform, or dentiform toward 
the apex, the longest cilia % to l /z as long as the width of the blade; apex long- 
acuminate, flat to slightly rounded, plane to slightly beveled in profile; setae arising 
evenly in form and often in color, % to nearly l / 2 as long as the blade, slightly to 
strongly attenuate, whitish-lutescent, translucent to subopaque. Sporophylls with 
the margins eciliate to short- cilia te toward the base; apex broadly rounded to 
carinate. Megaspores rather finely rugose-reticulate on the commissural face, more 
coarsely and less prominently marked on the outer face, pale orange-yellow. 

S. Wightii is the most generalized Old World species in the series and probably 
represents the basic type for that region as S. Sellowii and S. Sartorii do for Tropical 
America. It is most closely related to S, njamnjamensis. 

The African and Indian plants are very closely related. On the basis of rela- 
tively few collections it seems to be distinct but not in characters oi specific value. 
Ample material may indicate a need to change the varietal status given the African 
plant here. 

Southern India and Ceylon, eastern Africa. 


a. Setae predominantly strongly attenuate; sporophylls eciliate to only slightly 
ciliate toward the base. 10a. var. Wightii , p. 39 

a. Setae stout to predominantly slightly attenuate; sporophylls predominantly 
short-ciliate toward the base - — -10b. var. Phillipsiana, p. 40 

10a. Selaginella Wighto var. Wightii. Fig. 13. Map 19 

iginella Wightii var. ve 
Mauritius, Sieber ? B). 

1900, ex char. (Holotypc 

The localities taken from the literature (Map 19) are from Alston, in Proc 
Nat. Inst. Sci. India 11:215. 1945. Although I have not seen the type of var 

[Vol. 42 


vetusta said, probably in error, to come from Mauritius, the description does not 
adequately separate it from var. Wightii. In addition, I have seen Gamble 17258 
(P) identified by Hieronymus as var. vetusta and it is typical var. Wightii. 

Rocky places, 700-1700 m. 
Southern India and Ceylon. 
Specimens examined: 

India: Gamble 17258 (P); Wallich 2188/3 ( BM > US); Peninsula Ind. orientalis, 


us on 211 (US): Thwaites I did (P): W 

10b. Selaginella Wightii var. Phillipsiana Hieron. in Hedwigia 39:320. 
1900. (Holotype: Phillips B!). 

Selaginella Phillipsiana (Hieron.) Alston, in Jour. Bot. 77:222. 1939. 

The localities taken from the literature (Map 19) are from Alston, in Jour. 
Bot. 77:222. 1939. 

On stones and rocks, 1200-1500 m. 
British Somaliland to Tanganyika. 
Representative specimens: 

British Somaliland: /. B. Gillett 4681 (K), 4681 A (K); Glover & Gilliland 480 

(BM), 771 (BM); Godman 6 (BM) ; Darra-as, 5000 ft., June, 1895, Mrs. Lort Phillips 

Kenya: Napier 2054 (&)• 
Tanganyika: Uhlig 845 (BM). 

11. Selaginella cinerascens A. A. Eaton, in Fern Bull. 7:33. 1899. (Holo- 

Fig. 14. Map 20. 

Selaginella bryoides Underw. Our Nat. Ferns, ed. 6, 13 8. 1900, not (Kaulf.) Hieron., 

type: Kimball GH!). 


1901, not Lycopodium bryoides Nutt. ex Baker, 1887; considered as based on Selag- 
inella cinerascens A. A. Eaton. 

Stems with the older primary branches once to usually twice to rarely three 

Leaves herbaceous to 

herbaceous-papyraceous, none red, linear, acuminate to broadly lanceolate; base 
abruptly adnate, distinct from the stem in color or sometimes rather decurrent on 
the leader stem, usually glabrous, sometimes pubescent; margins rarely eciliate, 
usually ciliate, the cilia usually piliform, or dentiform only toward the apex, the 
longest cilia up to % as long as the width of the blade; apex acute to nearly obtuse, 
flat to slightly rounded, plane in profile; setae absent, the apex not or slightly 
modified, or if present, arising rather abruptly in form and col 
the blade, stout, whitish, opaque to translucent. Sporophy 



Megaspores slightly rugose on the commissural face, more prominently marked on 
the outer face, with an equatorial ring, pale yellow. 



S. cinerascens is quite closely related to the next species, S. Arsenei. The two 
may be distinguished by the cilia which are predominantly piliform in S. cinerascens 
and dentiform in S. Arsenei. There is a close resemblance to S. nivea of Madagascar 
but this does not extend beyond characters of habit and general aspect. 

This species affords the best illustration of the difference in length of leaves on 
the stem and branches, a character presumably due to some type of inhibition of 
growth. The leaves on the main stem are the longest, those on the primary branches 
are shorter and those on the secondary branches are the shortest. 

S. cinerascens is one of the few species that is not closely associated with rocky 
habitats. It grows on hillsides and slopes, among or beneath shrubs and small trees, 
often in clay soil, sometimes in rocky soil, from sea level to about 200 m. 

California and Baja California. 

Representative specimens: 

United States. California: Abrams 3399 (F, GH, MO, NY, US); April 12, 1918, 
Bethel (GH, MO, US) ; National City (San Diego), May, 1897, L. F. Kimball (GH, NY) ; 
Parish 10765 (NY, US). 

Mexico, baja California: Ferris 8474 (NY, US); Wiggins 4213 (GH, NY, US). 

12. Selaginella Arsenei Weath. in Jour. Arm Arb. 25:417. 1944. (Holotype: 
Arsene 10641 US! fragment GH!. Paratypes: Arsene 9983 US!, 10643 US!). 

Fig. 15. Map 21. 

Stems with the older primary branches once to twice pinnate; leafy stems 
radially symmetrical. Leaves herbaceous, none red, ligulate to ligulate-long- 
triangular; base abruptly adnate, distinct from the stem in color, glabrous to 
pubescent; margins ciliate, the cilia dentiform, the longest cilia about l /% as long 
as the width of the blade; apex acute to acuminate, rounded, plane or nearly so in 
profile; setae absent, the apex slightly modified, or if present, arising evenly in 
form and usually in color, about % as long as the blade, stout, greenish- to 
lutescent-whitish, subopaque. Sporophylls with the margins very short-ciliate 
toward the base; apex broadly rounded. Megaspores slightly and finely rugose- 
reticulate, lemon-yellow. 

The differences between S. Arsenei and the closely related S. cinerascens are 

discussed under S. cinerascens. 

S. Arsenei grows in rocky places, at one locality at 2700 m. 

Central Mexico. 
Specimens examined: 

Mexico, san luis potosi: T. C. # E. M. Frye 2983 (F). queretaro: Queretaro, 
1914, Arsene 10641 (GH, US), 9983 (US), 10643 (US), hidalgo: Hitchcock & Stan- 
ford 7248 (US). Guerrero: T. C. & E. M. Frye 3149 (US). 

13. Selaginella macrathera Weath. in Jour. Arn. Arb. 24:326. 1943. (Holo- 


Fig. 16. Map 22 

stems snort, witn tne oiaer primary orancucs unwc i-u **xw 7 
leafy stems radially symmetrical. Leaves herbaceous-papyraceous 


[Vol. 42 


ulate; base abruptly adnate, distinct from the stem in color, glabrous to slightly 
pubescent; margins ciliate, the cilia dentiform, or short-piliform only at the base, 
the longest cilia % or less as long as the width of the blade; apex acuminate to 
bluntly acuminate, broadly rounded, plane or nearly so in profile; setae arising 
abruptly in form and color, % to usually % as long as the blade, filiform, white or 
whitish, subopaque to translucent. Sporophylls with the margins short-ciliate 
toward the base; apex broadly rounded to carinate. Megaspores slightly to mod- 
erately finely rugose on the commissural face, moderately rugose on the outer face, 





Although known from only a single collection there is no doubt that this is a 
valid species. The dentiform cilia of the leaves and the very long setae distinguish 
it from all others. 

Collected on ledges on north-facing lava cliffs in Chihuahua, Mexico. 
Specimens examined: 

Mexico, chihuahua: Sierra del Virulento, 2-3 miles east of Rancho Virulento, ca. 
65 miles south of Ojinaga, Aug. 11, 1941, J. M. Johnston 8067 (GH, MO). 



14, Selaginella shakotanensis (Franch. ex Takeda) Miyabe & Kudo, in Jour. 
Fac. Agric. Hokkaido Imp. Univ. 26 (Fl. Hokkaido & Saghal. I):64. 1930. 

Fig. 17. Map 23. 

Selaginella rupestris (L.) Spring var. shakotanensis Franch. ex Takeda, in Bot. Mag. Tokyo 
23:237. 1909. (Holotype: Faurie 9895 SAP). 

Stems with the older primary branches once to twice pinnate; leafy stems some- 
what dorsiventral in the position of the leaves. Leaves fleshy to fleshy herbaceous, 
none red, ligulate to ligulate-lanceolate; base abruptly adnate, distinct from the 
stem in color, or sometimes rather decurrent on the leader stems, glabrous; margins 
ciliate, the cilia piliform, or dentiform only toward the apex, the longest cilia % 
to % as long as the width of the blade; apex convexly acute to obtuse, rounded to 
carinate, abruptly beveled to truncate in profile; setae arising abruptly in form 
and color, about */5 as long as the blade, attenuate, whitish, translucent. Sporophylls 
with the margins eciliate to long-ciliate toward the base; apex narrowly carinate. 
Megaspores finely and very slightly rugose-reticulate, orange. 

S. shakotanensis may be distinguished from S. sibirica of Rupestres, which 
overlaps its range, by the shorter setae and the abruptly adnate leaf-bases. The 
setae are about % as long as the blade in this species while they are % as long or 
longer in S. sibirica. The main stem may have some of the leaves with a decurrent 
base but on the branches the leaf-base is abruptly adnate. S. shakotanensis bears a 
rather close resemblance to S. mutica var. limitanea from which it may be sep- 
arated by its slightly dorsiventral leafy stem. 

Rocky alpine habitats; one collection was made at 2000 m. 


Specimens examined: 

Japan: Aug. 20, 1898, Faurie (P), 1 577 (P), Shakotan, Yezo Island, Shiribeshi, June 
9, 1893, Faurie 9895 (P, MO), 10543 (P, MO), 13151 (P, MO) 5 U. S. Nat. Herb, sheet 

15. Selaginella Wallacei 

1900, maintained over 

S. montanensis by Broun, Index N. Am. Ferns, 159. 1938. (Holotype: Walla 

B! fragment NY!). 

Fig. 18. Map 24. 

Selaginella montanensis Hieron. in Hedwigia 39:293. 1900. (Holotype: A. tf A. Krause 

B! fragment NY!). 
Selaginella rupestris (L.) Spring f. Wallaces (Hieron.) Clute, in Fern Bull. 16:52. 1908. 
Selaginella rupestris f. montanensis (Hieron.) Clute, in Fern Bull. 16:52. 1908, as 

Selaginella rupestris var. columbiana M. E. Jones, in Univ. Mont. Bull. Biol. Ser.- 15:8. 



Selaginella rupestris var. Wallacei (Hieron.) Frye, Ferns Northwest, 33. 1934. 
Selaginella RosendahlH Hieron, ex Frye, Ferns Northwest, 33. 1934, in synon. (Evidently 

based on Rosendahl 864). 
Selaginella Wallacei f. columbiana (M. E. Jones) Broun, Index N. Am. Ferns, 159. 1938. 
Selaginella rupestris var. montanensis (Hieron.) [incorrectly attributed to Frye by") 

Broun, Index N. Am. Ferns, 159. 1938, in synon. 

[Vol. 42 

btems with the older primary branches once to twice, rarely three times pinnate; 
leafy stems radially symmetrical to rather dorsiventral in position, length and shape 
of the upper and under leaves. Leaves herbaceous to fleshy, none red, ligulate to 
ligulate-long- triangular; base abruptly adnate, distinct from the stem in color, 
sometimes rather decurrent on the leader stems, glabrous or pubescent; margins 
eciliate or ciliate, the cilia dentiform to usually piliform, the longest cilia Y 5 to l / 4 
as long as the width of the blade; apex acute to obtuse, broadly rounded, abruptly 
beveled to truncate in profile, or predominantly so; setae arising abruptly in form 
and color, % to %, rarely %, as long as the blade, rather stout, milk-white and 
opaque to whitish or greenish-white and translucent. Sporophylls with the margins 
eciliate to short-cilia te toward the base; apex broadly rounded to subcarinate. 
Megaspores rugose to rugose-reticulate, usually with an equatorial ring, pale orange. 


It is 


and to a considerable extent its variability is correlated with its ecology. Specimens 
from very damp, shady situations have long stems that form loose mats and the 
leaves are somewhat distant. Those from dry, exposed rock or gravelly turf have 
short stems that form a compact mat and the leaves are close. 

Of the species that grow within its range it sometimes resembles S. densa var. 
scopulorum of Rupestres. It may be separated by the remote branches, S. densa 
having approximate ones, and by its abruptly adnate rather than decurrent leaf- 





inella Wallacei 

gravelly soil, or on moist, shaded rocks, bluffs and cliffs, in crevices or covering 
various types of igneous and sedimentary rocks, from sea level to 2000 m. 

Western Montana to southern British Columbia and California. 

Representative specimens: 

Canada. British Columbia: Hitchcock & Martin 7367 (MO, NY) ; Macoun 86172 

<™' NY ' US >' 86 3? 6 < GH ' NY » US), 86377 (GH, NY, US); Rosendabl 864 (BM, 
MIN, MO, NY, US). 

United States. Montana: Aug. 13, 1908, M. E. Jones (BM, MO, US); "Crossing", 
Northern Pacific Railroad, Oct. I, 1882, A. tf A. Krause (B, NY), idaho: Kirkwood & 
Severy 1718 (GH, US); H. T. Rogers 763 (GH, MO); Sandberg et al. 54 (GH, US). 
Washington: Constance fif Rollins 998 (F, GH, MO, NY, US); Otis 1149 (GH MO 
NY, US) ; Suskdorf 8563 (GH, MO, US) ; Thompson 8747 (F, MO, NY) ; S. M. a? E B. 
Teller 748 (GH, MO, US), oregon: Thompson 4002 (MO, NY, US), 4020 (MO NY 
VS),4W2 (MO, US); Oregon, Capt. Wallace (B, NY, YU); Wilkes Exped. "Wash. & 

?™ ?/A x ( £ H ' t ? Y ' US ,', YU) • CALIFORNIA: Baker 2746 (GH, NY, US) ; Beetle 1566 
(GH, MO, NY, US); Heller 5024 (GH, MO, NY, US). 

16. Selaginella mutica D. C. Eaton ex Underw. in Bull. Torr. Bot. Club 
25:128. 1898. (Lectotype by Weath. in Jour. Am. Arb. 25:413. 1944: 
Meehan YU! fragment NY!. Paratypes: M. A. C. Uvermore NY!; A. G. 
Compton NY!; E. J. Spence NY!; Tourney on Aug. 28, 1894 NY!; Brandegee 



NY!. Bigelow NY! and Mex. Bound. Survey NY! are excluded as types; they 
are intermediates between var. limitanea and var. mutica) . 

Stems with the older primary branches once to twice pinnate; leafy stems 
radially symmetrical. Leaves fleshy to fleshy-herbaceous, none red, ligulate- 
lanceolate to lanceolate-elliptical to lanceolate-ovate; base abruptly adnate, distinct 
from the stem in color, or slightly decurrent on the leader stems, glabrous or 
pubescent; margins ciliate, the cilia entirely piliform, the longest cilia l /± to % as 
long as the width of the blade, to entirely dentiform and shorter; apex obtuse to 
convexly acute, strongly carina te, subtruncate in profile; setae absent, the 
apex being scarcely modified, or if present, arising abruptly in form and usually in 
color, up to % as long as the blade, stout, whitish to greenish-white, subopaque to 
translucent. Sporophylls with the margins eciliate to long-ciliate toward the base; 
apex carinate to narrowly carinate. Megaspores rugose-reticulate to coarsely rugose 
on the commissural face, less marked to smooth on the outer face, pale orange. 

The variation of S. mutica, as pointed out by C. A. Weatherby, 23 is rather 
parallel to that of S. Underwoodii which has a very similar range. However, in 
S. mutica the material of Texas, southern New Mexico and southern Arizona is 
more distinctive and the line of demarcation stronger than in S. Underwoodii, so 
that I am recognizing two varieties. The characters of seta length and ciliation of 
the sporophylls correlate to a high degree and separate the major components of the 
species. The southern var. limitanea is itself variable but the extremes, one with 
short, dentiform, ascending cilia on the leaves, and the other with long, piliform, 
patent cilia, are too often connected by intermediates and the characters vary 
considerably in some individual mats. 

The muticous incurved leaves of var. mutica give the stems a reptilian aspect 
under low magnification. 

In Colorado I have seen S. mutica growing with S. Underwoodii in the same 
niche. From a number of such cases it appeared quite clear that although the mats 
of the two species were originally nearly equal in size, S. mutica gradually occupied 
more and more of the niche and finally entirely replaced S. Underwoodii. 

Rarely a strobilus may be slightly dorsiventral. 
Colorado and Utah to Texas and Arizona. 


a. Leaves setate; sporophylls with the cilia mostly or entirely ascending, all 
or some dentiform 16a - var - limitanea, p. 46 

a. Leaves muticous or nearly so; sporophylls with the cilia entirely or pre- 
dominantly spreading, piliform 16b. var. mutica, p. 46 


Jour. Arn. Arb. 25:415. 1944. 


[Vol. 42 



Weath. in J 

1944. (Holotype: Goodding US! fragment GH!. Paratypes: Ferriss GH!; 
Hinckley 1155 US!; Ingram 2723 US!; Moore 8 Steyermark 3046, in part, 
GH!; E. J. Palmer 31951 US!; Slater US!; Standley in 1906 US!; Wooton in 
1899 US!, in 1903 US!, in 1906 US!, in 1909 US!). Map 25. 

Selaginella mutica var. texana Weath. in Jour. Am. Arb. 25:414. 1944. (Holotype: 
Moore & Steyermark 31 96 GH! fragment US!. Paratypes: Moon? # Steyermark 3046, 
in part, GH!; E. J. Palmer 30871 GH!; Sperry 428 US!. Tracy & Earle 273 GH! US! 
and Mex. Bound. Survey YU fragment NY! are excluded as types; they are inter- 
mediate between var. limitanea and var. mutica) . 

Leaves short-setate; cilia mostly or entirely ascending, piliform to dentiform. 
Sporophylls with the cilia mostly or entirely ascending, all or some dentiform. 
Sheltered or moist cliffs and rocky hillsides, on igneous rocks, 1300-2400 m. 
Southwestern Texas to southeastern Arizona. 
Representative specimens: 

United States. 


Steyermark 3046 (GH, MO, US) ; ridge south of Emory Peak, Chisos Mountains, Brewster 
Co., 2300 m., June 22, 1931, Moore tf Steyermark 3196 (GH, MO, US); £. /. Palmer 
30871 (GH, MO, US), 31951 (MO, US); Sperry 428 (US), new Mexico: mountains 
south of Demmg, Oct. 4, 1937, Goodding (GH, US); Dec, 1924, Slater (US); Sept. 10, 
1899, Wooton (US), April 18, 1903 (NY, US), June, 1906 (US), Jan. 9, 1909 (US). 
Arizona: March, 1904, Ferriss (GH). 



Fig. 19. Map 26. 



Leaves muticous or nearly so; cilia entirely or predominantly spreading, pili- 
form. Sporophylls with the cilia entirely or predominantly spreading, piliform. 

The following specimens are intermediate between var. mutica and 


limitanea: Tracy tf Earle 275 (BM, F, GH, MO, NY, US) ; Mex. Bound. Survey, 


Exposed or sheltered rocky bluffs, cliffs and ledges, in crevices or soil pockets 

or usually covering rocks, on igneous rocks, sandstone or rarely limestone, 1450- 
4300 m. 

Colorado and Utah to Texas and Arizona 
Representative specimens: 

United States. Colorado: 1874-1878, Bn 


Compton (NY) 

x^t w^ 1, T ' MeehaU (NY ' YU) ; E ' h SpenCe (NY) ; K - M - d A - F > Try™ 507 1 (CU, 


(ARIZ DS, FI, K, MICH, MO, RM) ; E. P. Walker 365 (GH, NY, US), 'ut'ah: Graham 
7592 ( US ) , 10004 ( US ) ; Maguire 18450 ( GH ) . 


- o» 5083 (BM, GH, MO, US) ; Waterfall 4811 (GH, MO, NY), 

W 2 ri GH ' M °' NY) ; Wri * bt 2I15 (GH ' NY >- NEW Mexico: Arsene & Benedict 16643 
(F, US); Drouet e? Richards 3311 (F, GH, MO, NY, US); Standley 5199 (MO, NY, 
US). Arizona: Harrow 3631 (GH); 1896, Tourney (NY, US), 1894, Tourney CNY). 









17. Selaginelxa Extensa Underw. in Bull. Torr. Bot. Club 25:131. 1898. 
(Holotype: Pringle 3QOO NY. Paratype: C. Mullet NY). Fig. 20. Map 27. 

Selaginella rupestris mexicana Conzatti, Fl. Tax. Mex. 1:150. 1939, without Latin descrip- 
tion, not var. mexicana Milde, 1867. (Based on Pringle 3900). 

Plant terrestrial or sometimes epiphytic. Stems very long to long, branches 
long, remote, with the older primary branches once to twice pinnate; leafy stems 
radially symmetrical or nearly so. Leaves herbaceous, none red, subulate to ligu- 
late-long-triangular to ligulate-lanceolate; base abruptly adnate, distinct from the 
stem in color, usually glabrous, sometimes pubescent; margins cilia te, the cilia 
dentiform, the longest cilia less than l / 4 as long as the width of the blade; apical 
portion usually revolute; apex acuminate, rounded, plane or nearly so in profile; 
muticous, the apex subulate, greenish-hyaline. Sporophylls with the margins 
eciliate to short-ciliate toward the base; apex carina te. Megaspores finely 
reticulate on the commissural face, the equatorial region strongly and irregularly 
papillate, granular on the outer face, pale orange. 

S. extensa is casually epiphytic and rather similar in habit to the pronounced 
epiphyte S. oregana of Rupestres. The revolute apical portion of many of the 
leaves, the muticous essentially unmodified leaf-apex, the habit and the unusual 


megaspores combine to make this a very 


870-1700 m. 

[Vol. 42 


Eastern Mexico and Jalisco. 
Specimens examined: 

Mexico, tamaulipas: Sharp et al. 50255 (MO, US), san luis potosi: McVaugh 
10446 (US); Pennell 1/94/ (US); Las Canoas, Aug. 21, 1891, Pringle 3900 (F, GH, K, 
MO, US); Sharp 46264 (US). Veracruz: Bourgeau 2541 (GH, YU) ; Mohr 12 (YU); 
Purpus 6052 (F, GH, MO, US); Sharp 44168 (US), hidalgo: T. C. 6 E. M. Frye 2537 
(NY, US) ; Hoogstraal tf Chase 7308 (F, MO, US) ; Kenoyer & Crum 3910 (GH) ; Sharp 
45864 (US), puebla: Sharp 45375 (US), jalisco: McVaugh 11829 (US). 

18. Selaginella Wrightii Hieron. in Hedwigia 39:298. 1900. (Holotype: 

Wright 828 m). 

Fig. 21. Map 28 

Selaginella rupestris (L.) Spring f. Wrightii (Hieron.) Clute, in Fern Bull. 16:52. 1908. 

Stems with the older primary branches once to twice pinnate; leafy stems 
usually strongly to rarely moderately dorsiventral in position, length and shape of 
the upper and under leaves. Upper leaves ligulate-long-triangular to long-tri- 
angular; apex acuminate to convexly acute. Under leaves herbaceous, none red, 
longer to only slightly longer than the upper, linear; apex acuminate, flat to 
slightly rounded, plane or nearly so in profile. All leaves with the base abruptly 
adnate, distinct from the stem in color, or the under leaves with the base rather 
decurrent, pubescent or glabrous; margins cilia te, the cilia piliform, or dentiform 
only toward the apex, the longest cilia % to l / 2 as long as the width of the blade; 
setae absent, the apex being hardly modified, or if present, arising very evenly in 
form and usually in color, rarely up to % as long as the blade, stout, whitish- 
lutescent to lutescent, translucent to opaque. Sporophylls with the margins long- 
cihate toward the base; apex broadly rounded to carinate. Megaspores prominently 
rugose, pale to bright orange. 

ventral. The 


ferentiated setae are the primary characters of the species. This is the only species 
for which I have sufficient information to class definitely as a calciphile. 

Exposed or shaded ledges and pockets of limestone, 800-2300 m. 
Texas and New Mexico to southern Mexico. 
Representative specimens: 

United States, texas: E. J. Palmer 11389 (GH, MO, NY, US) ; Rose & Fitch 17973 
(MO, NY, US) ; Tbarp 43.514 (MO, NY, US) ; R. M. & A. F. Tryon 5034 (BM, GH, 
MO, P, US), 5035 (B, BM, F, GH, MO, NY, P, UC, US), 5037 (ARIZ, FI, GH, K, 
MO, RM, US); hills, Turkey Creek (near Cline, Uvalde Co.), June 25, 1849, Wright 
828 (B, GH, NY, US), new mexico: Standley 40389 (GH, US). 

Mexico, tamaulipas: Bartlett 10561 (F, US), nuevo leon: Barkley 16144U 
(GH, MO); Harvey 988 (GH, US); C. H. Muller 2638 (GH, US), coahuila: John- 
ston 8607 (GH, MO), 8854 (GH), 9106 (GH, MO, US), chihuahua: Harvey 1 507 
(GH, US) ; E. Palmer 455 in 188 5 (GH, US, YU). san luis potosi: A. A. Lundell 



9678 (G 

san luis potosi: A. A. Lundell 96 
kwood (or Kirkwood) 20 (MO, US), 142 (GH, MO). 
puebla: E. B. Copeland 1 74 (US). Guerrero: Hinton 
le & Conzatti 1390 (GH). 



19. Selaginella Steyermarkii Alston, in Ann. & Mag. Nat. Hist., XII, 7:638. 
1954. (Holotype: Steyermark 50 50 1 BM!. Para types: Steyermark 471 1 5 

BM!, 5I2Q2?M\). 

Figs. 22, 23. Map 29. 

Stems with the older primary branches once to twice pinnate; leafy stems 
strongly dorsiventral in position, length and shape of the upper and under leaves. 
Upper leaves subulate to subulate-long-triangular; apex long-acuminate. Under 
leaves papyraceous to herbaceous-papyraceous, usually some red, longer than the 
upper, subulate-lanceolate to ligulate-lanceolate; apex long-acuminate, flat or very 
slightly rounded, plane in profile. All leaves with the base abruptly adnate, distinct 
from the stem in color, pubescent to rarely glabrous; margins ciliate, cilia piliform, 

% to y 2 

% to y a 

strongly attenuate, white to whitish-lutescent, translucent to subopaque. Sporo- 
phylls with the margins eciliate to short-ciliate toward the base; apex rounded to 
carinate. Megaspores finely to moderately rugose-reticulate, to nearly smooth on 
the outer face, more prominently and coarsely marked in the equatorial region, 
pale orange. 

This species and the next, S. Hansenii, are the only ones that usually have some 
red leaves on the stem. In S. Sartorii such leaves are present infrequently. They 
tend to occur particularly among the oldest living leaves and the youngest dead 
ones and often are more common on the under side of the stem. 

The long-acuminate apex of the upper leaves and the thin under leaves afford 
an ample separation from S. Hansenii, This species has the most robust leafy stems 
of any in the series and material collected in the growing state presents particularly 
attractive sprays. It is one of the very few species I know to be casually culti- 
vated; the Standley collection cited below was from the garden of Mariano Pacheco. 

Rocks and banks, 1630-2000 m. 

Guatemala and adjacent Mexico. 

Specimens examined: 

Mexico, chiapas: Matuda 1846 (US). 

Guatemala: Aguilar 1429 (F, US); Hatch tf Wilson 283 (US), 325 (US); Haupt 
J OS (US) ; Maxon d Hay 3410 (US) ; Standley 63102 (F, US) ; Steyermark 47115 (BM) , 
between San Sebastian Hacienda and large penasco above town, Dept, Huehuetenango, 
2000-2200 m., 1942, Steyermark 50501 (BM), Steyermark 51292 (BM); Tuerckheim 
8844 (GH, US); L. O. Williams 14299 (US). 

20. Selaginella Hansenii Hieron. in Hedwigia 39:301. 1900, as Hanseni; 
maintained over S. Bolanderi. (Lectotype: Hansen 878 B!. Paratypes: Hille- 

brand B!; A. A. Eaton B!). 

Fig. 24. Map 30. 

Selaginella Bolanderi Hieron. in Hedwigia 39:300. 1900. (Holotype: Bolander 4511 B! 

fragment NY!). 
Selaginella rupestris (L.) Spring f. Hansenii (Hieron.) Clute, in Fern Bull. 16:52. 1908. 
Selaginella rupestris f. Bolanderi (Hieron.) Clute, in Fern Bull. 16:52. 1908. 
Selaginella rupestris var. Hansenii (Hieron.) Jepson, Man. PI. Calif. 42. 1923. 
Selaginella rupestris var. Bolanderi (Hieron.) Jepson, Man. PI. Calif. 42. 1923. 

[Vol. 42 


Stems with the older primary branches once to usually twice to rarely three 
times pinnate; leafy stems strongly dorsiventral in position, length and shape of 
the upper and under leaves. Upper leaves long- triangular to ligulate to lanceolate- 
long- triangular; apex convexly acute to acute. Under leaves herbaceous, usually 
some red, longer than the upper, linear-lanceolate to lanceolate; apex convexly 
acute to acuminate, flat to slightly rounded, plane to gently beveled in profile. All 
leaves with the base abruptly adnate, distinct from the stem in color, glabrous or 
pubescent; margins ciliate, the cilia piliform, or dentiform only toward the apex, 
the longest cilia l / 4 to rarely nearly l / 2 as long as the width of the blade; setae 
arising rather abruptly in form and in color, % to nearly J4 as long as the blade, 
stout to attenuate, milk-white to whitish, translucent to opaque. Sporophylls 
with the margins eciliate to short-cilia te toward the base; apex broadly rounded to 
carinate. Megaspores rugose to rugose-reticulate on the commissural face, less 
prominently rugose-reticulate to smooth on the outer face, with an equatorial ring, 


S. Hansenii is related to the previous species, S. Steyermarkii, by the dorsiventral 
leafy stem and the commonly occurring red leaves. These characters, however, may 
be the result of parallel evolution. The red leaves tend to be more common on 
the under side of the stem and in the area where living and dead leaves adjoin. 

The branches at the center of a mat tend to be rather erect and compact, those 
on the periphery prostrate, long-creeping and remote. This variation within a mat, 
together with its associated characters of number and spacing of the leaves, is 
similar to that found in S. tortipila of Arenicolae. 

Open or lightly shaded bluffs or cliffs and surfaces of igneous rocks, 330— 
1350 m. 


Central California. 

Representative specimens : 

United States, californ 


(B, MO, NY, US); 1892, A. A. Eaton (B, NY); Fisher's Cabin, Mokelumne River, 
Amador and Calaveras counties, April, 1893, G. Hansen 878 (B, MO, NY, US); Heller 
2583 (F, GH, MO, NY, US), 5466 (GH, MO, NY, US), 8173 (F, GH, MO, NY, US), 
10715 (GH, MO, US), 11802 (F, GH, MO, US); 1863, Hillebrand (B); R. M. (g A. F. 
Tryon 5067 (ARIZ, B, BM, CU, DS, F, FI, GH, K, MICH, MIN, MO, NY, P, POM, 
RM.UC, US, WS). 

21. Selaginella carinata Tryon, spec. nov. Figs. 25, 26. Map 31. 

Caules breves, ramis brevibus propinquis; caules frondosae positione foliorum 
superorum et inferorum valde dorsiventrales sed forma minus. Folia carnosa, apice 
carinato vel valde carinato, aspectu laterali truncato vel abrupte declivi, base 
abrupte adnata cum caule distincta colore, setis abrupte productis colore sensim 
diversis. Sporophylla marginibus ad basem eciliatis. Megasporae laeves flavae. 
Typus: Rose, Pachano & Rose 2303Q (US). 

Stems short; branches short, approximate; older primary branches once to 
usually twice, sometimes three times pinnate; leafy stems strongly dorsiventral in 
position and less so in the shape of the upper and under leaves. Upper leaves 




ligulate-long- triangular; apex convexly acute. Under leaves fleshy, none red, 
lanceolate to ligulate-lanceolate; apex acuminate to convexly acute, carinate to 
strongly carinate, truncate to abruptly beveled in profile. All leaves with the base 
abruptly adnate, distinct from the stem in color, glabrous; margins cilia te, the cilia 
piliform toward the base, dentiform above, the longest cilia % or less as long as 
the width of the blade; setae arising abruptly in form and evenly in color, ]/ 4 to l /z 
as long as the blade, rather stout at the base, tapering evenly to the tip, greenish- 
to lutescent-whitish and subopaque at the base, milk-white and opaque toward the 
tip. Sporophylls with the margins eciliate toward the base; apex strongly carinate. 
Megaspores smooth, yellow. 

S. carinata is a distinctive species and not evidently related to any others. The 
very short stems and branches and the fleshy leaves with carinate and truncate 
apex set it off from the other species of this series. Some strobili are slightly dorsi- 
ventral, particularly toward the base, apparently a development parallel to that 
in the African and Madagascar species. 

Central Ecuador. 

Specimens examined: 


'-W -' 

t • 


* / 


• ^ 

i fc* 

^* 'i 

/.-**. r^ 


m j^*r*jf ^W. * 

1 X * M 

• \ * ior\ X 

£}>M\ \j\ 


^T * 





WWN> »» *#•** 

.7^ * 

l r *y 



/ ■ 

v * 

«. f 


*» . ■ ■ 

"f '• 



* ■ _ ■ 





[Vol. 42 


22. Selaginella indica (Milde) Tryon, comb. nov. Fig. 27. Map 32. 

Selaginella rupestris (L.) Spring f. indica Milde, Fil. Europ. Atlant. 262. 1867. (Holo- 

ty pe : Hooker f. & Thomson ) . 
Selaginella longipila of authors, not Hieron., for example, of Alston, in Proc. Nat. Inst. 

Sci. India 11:214. 1945. 

Stems with the older primary branches once to twice pinnate; leafy stems 
strongly dorsiventral in position, length and shape of the upper and under leaves. 
Upper leaves subulate-long-triangular to long- triangular; apex acuminate. Under 
leaves papyraceous, none red, longer than the upper, long-triangular to ligulate- 
long- triangular to subulate-lanceolate; apex acuminate, flat, plane in profile. All 
leaves with the base abruptly adnate, distinct in color, glabrous to slightly pubes- 
cent; margins ciliate, the cilia piliform toward the base, dentiform above, the 
longest cilia */4 to l /z as long as the width of the blade; setae usually arising evenly 
in form, less so in color, % as long as the blade, strongly attenuate, tawny to 
lutescent-whitish, translucent. Sporophylls with the margins eciliate to short- 
ciliate toward the base; apex rounded to carinate. Megaspores rather finely rugose- 
reticulate on the commissural face, less prominently marked on the outer face, pale 

This species was known as S. longipila from the time of the description of that 
species by Hieronymus, since the specimens were cited as from the Himalayas. 
However, the labels on Hieronymus' material were erroneous, the specimens being 



its dorsiventral leafy stem and thin under leaves. The localities taken from the 
literature (Map 32) are from Alston, in Proc. Nat. Inst. Sci. India 11:214. 1945. 

Cliffs, rocky banks and ledges, 700-2800 m. 

Nepal, Bhutan, northern and western India. 

Specimens examined: 

Nepal: Polunin 14.78 (BM). 

Bhutan: Griffith 2871 (BM) ; Ludlow et ai. 17039 (BM), 18606 (BM). 

India: Duthie 3727 (BM, US); Khasia, 5000 ft., /. D. Hooker tf T. Thomson (GH, 
NY, P, YU) . 

23. Selaginella njamnjamensis Hieron. in Hedwigia 39:312. 1900. (Holo- 


Fig. 28. Map 33 

Stems with the older primary branches once to usually twice pinnate; leafy 
stems dorsiventral in position, length and slightly in the shape of the upper and 
under leaves. Upper leaves lanceolate to long-triangular; apex acuminate. Under 
leaves herbaceous, none red, longer than the upper, linear-lanceolate to lanceolate to 
long-triangular; apex acuminate, flat to slightly rounded, plane in profile. All 
leaves with the base abruptly adnate, distinct from the stem in color, glabrous to 
pubescent; margins ciliate, the cilia usually piliform toward the base, dentiform 
above, the longest cilia l / 4 to nearly l / 2 as long as the width of the blade; setae 
arising evenly in form and color, ]/ 4 or less as lone as the blade, stout to attenuate, 



tawny to lutescent-whitish, subopaque to opaque. Sporophylls on the upper side 
of the strobilus lanceolate, acuminate, on the under side ovate-long-triangular 
to narrowly ovate, acuminate, slightly longer and broader; margins short-ciliate 
toward the base; apex broadly rounded to carina te. Megaspores granular to 
granular-tuberculate, yellow. 

This species is the least dorsiventral of the related species in Africa and Mada- 
gascar. Although sufficiently distinct from S. caffrorum, both species tend to 
overlap in the range of variation of the individual characters. The herbaceous 
under leaves with the base abruptly adnate and distinct in color and the poorly 
differentiated setae are the primary characters of S. njamnjamensis. 

On and between exposed rocks, 800-1200 m. 

Central and southeastern Africa. 

Specimens examined: 

Anglo-Egyptian Sudan: Dandy J/o (BM); Hoyle 520 (BM). 

35S7 (BM). 




24. Selaginella caffrorum (Milde) Hieron. in Hedwigia 39:313. 1900. 

Fig. 29. Map 34. 

Selaginella rupestris (L.) Spring f. caffrorum Milde, Fil. Europ. Atlant. 262. \%67. 
(Holotype: Terra Caffrorum, Bunge). 

Selaginella rupestris var. incurva A. Br. in Kuhn, Fil. Afr. 213. 1868, category taken 
from annotation of Welwitsch 169 by A. Br. (Lectotype: Drege, Lycopodium 
rupestre a B. Paratypes: Ecklon S? Zeyber 7 B; Drege, Lycopodium rupestre aa B; 
Weltuitsch 169 B!; Quar tin-Dillon; Steudner). 

Selaginella rupestris var. incurva f. abyssinica A. Br. in Kuhn, Fil. Afr. 213. 1868, nomen 
nudum. A single description was provided for "forma capensis et abyssinica". I have 
seen Quar tin -Dillon, P, but not Steudner, the other specimen cited. 

Selaginella rupestris var. incurva f. angolensis A. Br. in Kuhn, Fil. Afr. 213. 1868. (Holo- 
type : Welwitsch 169BI). 

Selaginella rupestris var. incurva f. capensis A. Br. in Kuhn, Fil. Afr. 213. 1868, nomen 

nudum (see f. abyssinica). 
Selaginella capensis Hieron. in Hedwigia 39:314. 1900, epithet from S. rupestris t capensis 
A. Br. (Lectotype: Drege, Lycopodium rupestre a B. Paratypes: Ecklon # Zeyher 7 
B; Drege, Lycopodium rupestre aa B; K. Baur IIIO B; Sonder B!; Goldschmid B; 

Kehmann 3924 B, 4005 B; Sutherland B). 
Selaginella Quartiniana Cufodontis, in Phyton 4:178. 1952, nomen nudum, based on S. 
rupestris var. incurva f. abyssinica A. Br., nomen nudum. 

Stems with the older primary branches once to usually twice to rarely three 
times pinnate; leafy stems dorsiventral in position, length and shape of the upper 
and under leaves. Upper leaves linear-long-triangular to ligulate-long- triangular; 
apex acuminate to convexly acute. Under leaves papyraceous to papyraceous- 
herbaceous, none red, longer than the upper, lanceolate to ligulate-lanceolate to 
triangular; apex acuminate to convexly acute, flat to slightly rounded, plane or 
nearly so in profile. All leaves with the base abruptly adnate, distinct from the 

[Vox.. 42 


stem in color, or the under leaves with the base slightly decurrent and blending in 
color, glabrous or rarely pubescent; margins cilia te, the cilia piliform, or sometimes 
dentiform toward the apex, the longest cilia 54 to J4 as ^ on S as t ^ ie width of the 
blade; setae arising rather abruptly in form and color, % to Yz a $ l° n g as the 
blade, rather stout, whitish to whitish-tawny or whitish-lutescent, translucent to 
subopaque. Sporophylls on the upper side of the strobilus ovate-lanceolate to long- 
triangular, on the under side lanceolate-ovate to broadly ovate, acuminate, longer 
and broader; margins short- to long-ciliate toward the base; apex rounded to 
carinate. Megaspores rugose-reticulate to granular-rugose on the commissural face, 
more finely and less prominently marked on the outer face, lemon-yellow. 

The thin under leaves and well-differentiated setae are characteristic of S. 
caffrorum and set it off from the related S. njamnjatnensis. In addition, S. caffrorum 
usually has some of the under leaves on the main stem with decurrent bases. 

Dry or moist rocky places, 600-2000 m. 

Anglo-Egyptian Sudan to Angola and Union of South Africa. 

Specimens examined: 

Anglo-Egyptian Sudan: Jackson 1116 (BM); MacLeay 114 (BM). 
Tanganyika: Greenway 2431 (BM). 

Angola: Antunes 343 (P) ; DekJndt 543 (K, P); Lopolo, Huilla, 5000 ft., April, 
1860, Welwitsch 169 (B, K, NY, P). 

Union of South Africa, natal: Kudatis 8 $7 (P). orange free state: Kehmann 
3924 (P), 3949 (NY, P). basutoland: Dieterlen iij (P). cape of good hope: T. 
Cooper 662 (NY) ; Katberg, 4000-5000 ft., Lycopodium rupestre a, Drege (GH, K, P, 
US) ; Ecklon ti Zeyher (B, P) ; Sander (B). 


25. Selaginella echinata Baker, in Jour. Linn. Soc. 22:536. 1887. (Holo- 

type: Baron 4226 K!). 

Figs. 30, 31. Map 35 


pinnate; leafy stems dorsiventral in position, length, shape and in setae characters 
of the upper and under leaves. Upper leaves ligulate-long-triangular; apex acute; 
setae arising abruptly, mostly soon deciduous, usually substellate, divided into two 
to several filiform branches, whitish, translucent to opaque. Under leaves her- 
baceous-papyraceous, none red, longer than the upper, ovate-lanceolate; apex 



opaque to subopaque. All leaves with the base adnate to abruptly adnate, distinct 
from the stem in color, glabrous; margins ciliate, the cilia piliform, the longest 

to usually x /% as long as the width of the blade. Sporophylls on the upper 



ovate-triangular, acuminate, longer and broader; margins long-ciliate toward the 
base; apex broadly rounded to carinate. Megaspores tuberculate, yellow. 

S. echinata is the most distinctive of the three species of Madagascar. It is the 
most strongly dorsiventral species and in addition has the truly unique character 
of substellate setae on the upper leaves. Also some leaves and sporophylls have 




a unique whitish to tawny, long-appressed pubescence on the back, which among 
the specimens I have seen is most pronounced in Perrier 8262. 

Madagascar, 500-2000 m. 

Specimens examined: 

Madagascar: Baron 4226 (BM, K) ; 1879, Cowan (BM) ; Humbert 2823 (P), 3000 
( p )> 3538 (P), 1 1778 (P); Humbert & Swingle 4791 (P); Perrier 1168 (P), 8261 

<P2, 8262 (BM, P), 8311 (BM, P), 8312 (P), 8313 (BM, P), 8324 (P), 8346 (BM, P). 

26. Selaginella nivea Alston, in Perrier, Cat. PL Madagascar, 71. Feb. 1932, 
nomen nudum; in Dansk Bot. Ark. 7 (C. Chr., Pterid. Madagascar) : 194. June, 

1932. (Holotype: Perrier 8303 BM!. Paratype. Perrier 8305 BM!). 

Fig. 32. Map 36. 

Stems with the older primary branches once to twice pinnate; leafy stems dorsi- 
ventral in position and shape of the upper and under leaves. Upper leaves linear- 
lanceolate to lanceolate; apex acute to acuminate. Under leaves papyraceous, none 
red, triangular to ligulate-triangular; apex acute to acuminate, flat to slightly 
rounded, plane in profile. All leaves with the base abruptly adnate, distinct from 
the stem in color, glabrous; margins cilia te, the cilia piliform toward the base, 



white, opaque. Sporophylls on the upper side of the strobilus lanceolate-ovate, 
on the under side broadly ovate, acuminate, longer and broader; margins long- 

[Vol. 42 


cilia te toward the base; apex broadly rounded to carina te. Megaspores finely 
granular to nearly smooth, light yellow-orange. 

This species has a remarkable superficial resemblance to S. cinerascens although 
in technical characters there is no relationship between them. S. nivea may be 
separated from the next species, S. proxima, by the straight and relatively short 
setae in addition to the characters of habit. S. nivea is closely prostrate with the 
branches once or twice divided. S. proxima has the primary branches tending to 
be ascendent and dendroid and commonly three, rarely four, times divided. 

Madagascar, among bushes. 

Specimens examined: 

Madagascar: Humbert 11612 (P) ; Humbert & Swingle 5518 (BM, GH, P, US); 
vicinity of Ampanihy, June, 1910, Perrier 8303 (BM, P), Perrier 8305 (BM, P), 18702 


27. Selaginella proxima Tryon, spec. nov. Figs. 33-35. Map. 37. 

Caules ramis primariis saepe dendroideis; caules frondosae positione longitudine 
et forma foliorum superorum et inferorum dorsiventrales. Folia base abrupte 
adnata caule distincta colore, marginibus ciliis longissimis longitudine %— l /z 
laminae latitudinis, setis %-J^z laminae longitudinis valde curvatis juventate 
praesertim cretaceis caecis. Strobili dorsiventrales sporophyllis stichorum duorum 
dorsalium deltoideis duorum ventralium ovato-deltoideis longioribus et latioribus. 
Megasporae leviter rugoso-reticulatae vel laeves citreo-flavae. Typus: Humbert d 
Swingle 5705 (US). 

Stems with the primary branches tending to be ascendent and dendroid, the 
older primary branches twice to usually three to rarely four times pinnate; leafy 
stem dorsiventral in position, length and shape of the upper and under leaves. 
Upper leaves lanceolate to ligulate-lanceolate; apex acute to obtuse. Under leaves 
herbaceous, none red, longer than the upper, triangular to ligulate-lanceolate; apex 
acute to obtuse, flat, plane in profile. All leaves with the base abruptly adnate, 
distinct from the stem in color, glabrous to pubescent; margins ciliate, cilia pili- 
form, the longest cilia % to ]/z as long as the width of the blade; setae arising 
evenly to abruptly in form, abruptly in color, % to l /i as long as the blade, stout 
at the base, usually with a filiform tip, strongly curved, especially those of the 
branch-tips, milk-white, opaque. Sporophylls on the upper side of the strobilus 
deltoid to long-deltoid, on the under side ovate-deltoid, longer and broader; margins 
short- to long-cilia te toward the base; apex broadly rounded to carina te. Mega- 
spores very slightly rugose-reticulate to smooth, lemon-yellow. 

The strongly curved setae suggest a relationship to S. Dregei while in other 
characters S. proxima is more closely related to the preceding species, S. nivea. 
The differences are discussed under that species. 

A bush xerophile, 20-1200 m. 


Specimens examined: t 



Madagascar: Humbert 6557 (P), 7089 bis (P), 12439 (P), 12706 (P), 13338 (P), 
141 58 (P) ; Humbert & Swingle 558 1 (P), vicinity of Fort Dauphin, near Bevilany, 200- 
300 m., Sept. 14, 1928, Humbert & Swingle 5705 (GH, P, US). 

28. Selaginella Dregei (Presl) Hieron. in Hedwigia 39:315. 1900. 

Fig. 36. Map 38. 


583, reprint 153. 1844. (Holotype: Drege, Lycopodium rupestre b PRC. Drege, 

Lycopodium rupestre a is excluded as a type; it is S. caffrorum). 
Selaginella rupestris (L.) Spring f. Dregei (Presl) Milde, Fil. Europ. Atlant. 262. 1867. 
Selaginella rupestris var. recurva A. Br. in Kuhn, Fil. Afr. 213. 1868, category taken from 

annotation of Welwitsch 48 and Drege, Lycopodium rupestre b by A. Br. (Lectotype: 

Drege, Lycopodium rupestre b B!. Para types: Welwitsch 48 B!, 49 B). 
Selaginella rupestris var. recurva f. Dregeana A. Br. in Kuhn, Fil. Afr. 214. 1868, based 

on Lycopodium Dregei Presl as to Drege, Lycopodium rupestre b. 
Selaginella rupestris var. recurva f. Welwitschiana A. Br. in Kuhn, Fil. Afr. 214. 1868; 

(Lectotype: Welwitsch 48 B!. Paratype: Welwitsch 49 B). 
Selaginella Dregei var. pretoriensis Hieron. in Hedwigia 39:317. 1900. (Lectotype: Reh- 

mann 4333 B!. Paratype: Wilms 1814 B). 
Selaginella Dregei var. Rehmanniana Hieron. in Hedwigia 39:317. 1900. (Lectotype: 

Rehmann 5576 B!. Paratype: Braga 98 B). 
Selaginella Dregei var. Bachmanniana Hieron. in Hedwigia 39:317. 1900. (Holotype: 

Bachmann 9 B ! ) . 
Selaginella Dregei var. Petersiana Hieron. in Hedwigia 39:317. 1900. (Holotype: Peters 

Selaginella Dregei var. Hildebrandtiana Hieron. in Hedwigia 39:317. 1900. (Lectotype: 
Hildebrandt 2363 B!. Paratypes: Fischer 627 B, fragment NY!; Stuhlmann 910 B, 

4305 B). 



Selaginella grisea Alston, in Jour. Bot. 77:222. 1939, based on S. Dregei var. Hilde- 
brandtiana Hieron. 

Stems often irregularly ascendent, buds rarely present at the base of the 
branches, older primary branches once to twice, rarely three times, pinnate; leafy 
stems radially symmetrical to somewhat dorsiventral in position and sometimes also 
in shape of the upper and under leaves. Leaves herbaceous to herbaceous-papy- 
raceous, none red, subulate to long-triangular to ligulate-long-triangular; base 
abruptly adnate, distinct from the stem in color, glabrous; margins ciliate, the 
cilia piliform, the longest cilia l /i as long to as long as the width of the blade; apex 
acuminate to long-acuminate, flat to slightly rounded, plane in profile; setae arising 
evenly in form and usually in color, % to usually l /z as long to as long as the blade, 
attenuate, usually strongly curved, especially those of the growing-points, milk- 
white and opaque to rarely tawny-whitish, subopaque. Sporophylls in 2 ranks on 
the under side of the branch-tip; margins eciliate to long-ciliate toward the base; 
apex broadly rounded. Megaspores slightly tuberculate on the commissural face, 
similar to slightly rugose on the outer face, with an equatorial ridge, yellow to 


The unilateral strobilus of this species is one of the most unique characters of 
any in the section. There are two rows of sporophylls on the under side of the 


[Vol. 42 

branch tip while vegetative leaves replace the usual two rows on the upper side. 
This character is sufficient to identify the species in spite of considerable variation 
in other characters such as the color, length and degree of curving of the setae, 
the leaf shape, the symmetry of the leafy stem, the number and length of the cilia 
on the leaves and the shape of the sporophylls. Except that the dorsiventral speci- 
mens have the broader under leaves, none of the variations correlate with each 
other or with a coherent distribution. 

Some sterile material with straight setae such as Peters, Mozambique, is difficult 
to place. However, such specimens as Moss l8gS8 BM which is otherwise similar, 
are fertile and can be placed with certainty. 

S. Dregei is variable in habit, the stems varying from closely prostrate to loosely 
and irregularly ascending. In the latter case the rhizophores may be unusually 
long. Rarely, some stems will have short branches at the base which will actively 
grow upon the death of the portions above. This character is otherwise found only 

and S. arenicola. 


The localities taken from the literature (Map 38) are from Alston, in Jour. 

Bot. 77:222. 1939. 

Exposed or sheltered rocky places, on or among igneous or sandstone rocks, 
700-2300 m. 

Southern and eastern Africa. 
Representative specimens: 

Uganda: Eggelmg 2619 (US). 

Kenya: Fischer 627 (B, NY) ; Egu, between Duruma and Taita, Jan. 1877, Hildebrandt 
2363 (B,BM,K,NY,P). 

Portugese East Africa (Mozambique): Gomes d Sousa 85 (BM); Peters (B). 
Rhodesia: Munro 1777 (BM, P, US), 1883 (BM); Rodin 4403 (MO, US). 
Angola: Exell d Mendonca 169 (BM); Pedras de Guinga, Pungo Andongo, 3800 ft., 
March, 1837, Welwitsch 48 (B, K, P). 
Bechuanaland: Burchell 2315 (GH). 
Union of South Africa. Transvaal: Leendertz 935 (BM, P), 2568 (F); Aapies- 

(NY, P). 


inaial: umsamcaDa, 1844, Lycopodmm rupestre b, Drege (B, K, P) ; W. T. 
Gerrard (P) ; /. M. Wood 1 1950 (F). orange free state: Kehmann 394.9 (P). cape 
of good hope: Pondoland, 1887-88, Bachmann 9 (B, P). 

Series Rupestres Tryon, ser. nov. 

Rhizomata et stolones nulla, gemmae breves simplices ad bases caulium rare 
praesentes vel absentes. Caules prostrati, ramis rare pendulis. Apices ramorum recti 
vel leviter curvati statu inerte. Caules frondosi leviter dorsiventrales vel radialiter 
symmetricales. Folia base decurrente vel valde decurrente. Typus: Selaginella 
rupestris (L.) Spring. 

Plants terrestrial or, in S. oregana, usually epiphytic; rhizomes and stolons 
absent, short, simple basal buds absent or, in S. oregana, occasionally present. 
Stems prostrate with rhizophores produced generally throughout or, in S. oregana, 



the branches usually pendent with rhizophores only on the prostrate main stems. 
Branches straight or slightly curled in the dormant state or, in S. oregana, strongly 
curled; branch tips straight or slightly curved in the dormant state. Leafy stems 
radially symmetrical, the leaves equal in position, length and shape on all sides on 
the same portion of the stem, or dorsiventral in position, the under leaves loosely 
appressed, the upper erect- ascending, and sometimes also in length, the under 
slightly longer than the upper; zone of green leaves equal or nearly equal on all 
sides of the stem. Leaf-bases strongly decurrent on all sides of the stem or those 
on the upper side decurrent. Setae of the sporophylls terete or oval at the base or 
often, in S. densa and S. sibirica, with the base strongly broadened and flattened. 
Megaspores 4 in a megasporangium; rarely 1-2, or in S. rupestris commonly 1-2. 
The Rupestres is a rather homogeneous series. Although the species are suf- 
ficiently distinctive they do not differ from each other by as many characters as 
the members of the other series. The last three species, S. utahensis, S. leucobryoides 
and S. asprella, share the peculiar character of easily fragmenting stems and this 
is also present, although to a lesser degree, in S. Watsonii. These four species form 
the only definite group within the series. S. Vardei, with some of the leaves rarely 
with abruptly adnate bases, may illustrate a transition from Arenicolae. Th 
dorsiventral S. densa appears to be an example of parallel evolution and not related 
by that character to Eremophilae or the dorsiventral Sartorii. S. rupestris is notable 
for its development of apogamy and S. densa and S. sibirica rarely show a tendency 
toward such a development. 


a. Dry leafy stems persistently whole, not fragmenting; stems forming 
cushion mats with discrete branches, or spreading mats with intricate 
branches, or pendent, forming festoons, or if forming cushion mats with 
intricate branches then the apex of the upper leaves fleshy, b. 
b. Stems forming flat mats with discrete branches or spreading mats with 
intricate branches, or pendent, forming festoons; apex of the upper 
leaves herbaceous to slightly fleshy, plane to abruptly beveled in profile; 
or if the apex of the upper leaves fleshy and truncate in profile then the 
setae of the leaves % to over Yz as long as the blade and the broadest 
sporophylls about 4 times as broad as the leaves, c. 
c Stems elongate, branches long and remote, intricate; apex of the 
upper leaves plane in profile; broadest sporophylls about 4 times as 

broad as the leaves. China 29. S. Vardei, p. 61 

c. Stems short, the branches short, approximate, discrete; or the apex of 
the upper leaves beveled to truncate in profile; or the broadest sporo- 
phylls about 2 times as broad as the leaves, d. 

d. Plants usually epiphytic, the stems pendent with rhizophores only 
at or near the base; leaves adnate to the stem for l / 4 to usually % 
to nearly l / 2 their length; branches strongly curled in the dormant 
state. Coastal Washington to northern California — 30. S. oregana, p. 61 

[Vol. 42 


d. Plants terrestrial, usually with rhizophores throughout; upper 
leaves adnate to the stem for % to rarely l / 4 their length; branches 
not or slightly curled in the dormant state, e. 
e. Apex of upper leaves nearly plane to abruptly beveled in profile, 
or if truncate then the setae lutescent and the stems forming 
compact flat mats with discrete branches, f. 
f , Upper and under leaves essentially or quite equal in length on 
the same portion of the stem; stems forming open, spreading 
mats with intricate branches; base of setae of the sporophylls 
not or slightly broadened and flattened, g. 
g. Broadest sporophylls about 2 times as broad as the leaves; 
sporophylls and leaves eciliate or with cilia strongly ascend- 
ing and dentiform toward the apex. Texas to Arizona, 

north to Wyoming 31. S. Underwoodii, p. 62 

g. Broadest sporophylls about 4 times as broad as the leaves; 
sporophylls and leaves with cilia spreading to laxly ascend- 
ing and piliform toward the apex. Georgia to Greenland, 

west to Arkansas, Nebraska and Alberta 32. S. rupestris, p. 64 

f . Upper and under leaves unequal in length on the same portion 
of the stem, the under definitely longer; stems forming com- 
pact mats with discrete branches; base of setae of the sporo- 
phylls often strongly broadened and flattened. Texas to 
Arizona; California; north to Saskatchewan, Alaska and 

British Columbia 33. S. densa, p. 66 

e. Apex of the upper leaves truncate in profile; setae white to 
tawny; stems forming open, spreading mats with intricate 

branches. Yukon to Siberia and Japan 34. S. sibirica, p. 71 

b. Stems forming rounded cushion mats with intricate branches; apex of 
the upper leaves fleshy, truncate or subtruncate in profile; setae of the 
leaves % to %, rarely %, as long as the blade; broadest sporophylls 
about 2 times as broad as the leaves. California to Oregon and 

Montana 35. S. Watsonii, p. 72 

a. Dry leafy stems readily fragmenting; stems forming cushion mats with 
usually intricate branches; apex of the upper leaves herbaceous to slightly 
fleshy. Southwestern Utah to southern California, h. 
h. Setae not forming a conspicuous tuft at the growing-tip, % or less as 
long as the blade, i. 

i. Setae essentially absent or, if present, then smooth, whitish to greenish- 
or lutescent-whitish, subopaque. Southwestern Utah and southeastern 

Nevada 36. S. utahensis, p. 74 

i. Setae scabrous, white, usually opaque. Southeastern California 

37. S. leucobryoidesy p. 74 



h. Setae forming a conspicuous tuft at the growing-tip, % to more than 
l /i as long as the blade, scabrous, white, usually translucent. Southern 
California 38. S. asprella, p. 75 

29. Selaginella Vardei Lev. Cat. PL Yun-Nan, 172. 1917. (Holotype: Moire 

56 E). 

Fig. 37. Map 39. 

Stems long, forming open, spreading mats; branches long, remote, intricate, not 
fragmenting when dry. Upper and under leaves about equal in length on the same 
portion of the stem; leaves subulate to subulate-long-triangular; base pubescent, 
that of the upper leaves adnate to the stem for % or less their length; margins 
cilia te, the cilia strongly ascending toward the apex of the blade; apex of the upper 
leaves herbaceous, flat to slightly rounded, plane or nearly so in profile; setae form- 
ing conspicuous tufts at the dry branch-tips, % to l /z as long as the blade, usually 
slightly scabrous, tawny to whitish-tawny, translucent. Sporophylls with the cilia 
dentiform and ascending toward the apex, the broadest about 4 times as broad as 
the leaves. Megaspores unknown. 

On the basis of its generalized characters, this species is placed as the most 
primitive in the series. One specimen examined had several of the leaves abruptly 
adnate and distinct in color at the base, characters of the previous series, Sartorii. 

In the material examined only microsporangia were found and these occurred 
in the basal sporophylls as well as those above. The explanation of this in terms of 
the reproduction of the species is not clear. It suggests that the stems may be 
monoecious or dioecious but fertilization would be extremely unlikely under such 
circumstances. The localities taken from the literature (map 39) are from Alston, 
in Bull. Fan Mem. Inst. 5:267. 1934. 

Exposed rocky places, 1500-3800 m. 

Szechwan, Yunnan and Tibet. 

Specimens examined: 

China, szechwan: H. Smith 2394 (BM); E. H. Wilson 5411 (P). yunnan: /. W. 
& C. /. Gregory (BM); Siao-ou-long, 2900 m., Oct. 1912, Moire 56 (BM). Tibet: Lud- 
low et al. 41 16 (BM), 5404 (BM), 1 4233 (BM). 


Wats. Bot. Calif. 2:350. 1880. 

Weath. in Jour. Arn. Arb. 25:411. 1944: Kautz YU!. Para- 

type: Summers 220Q YU!). 

Fig. 38. Map 40. 

Plants epiphytic with rhizophores only at the base of the stems, less often 
terrestrial with rhizophores borne throughout. Stems very long to long, forming 
pendent festoons; branches long, remote, intricate or discrete, strongly curled in 
the dormant state, not fragmenting when dry; or terrestrial stems forming an 
irregular mat with intricate branches. Upper and under leaves essentially or quite 
equal in length on the same portion of the stem; leaves long- triangular to ovate- 
triangular, excluding the adnate base; base glabrous, rarely pubescent, that of the 
upper leaves adnate to the stem for l / 4 to nearly l / 2 their length; margins eciliate 
to ascending-ciliate toward the apex; apex of the upper leaves slightly fleshy, 

[Vol. 42 


rounded to narrowly carinate, plane to gently beveled in profile; setae not or hardly 
forming conspicuous tufts at the dry branch- tips, about % as long as the blade, 
smooth, greenish, whitish or lutescent, translucent to rarely subopaque. Sporophylls 
. eciliate toward the apex, the broadest about 2 times as broad as the leaves. Mega- 
spores rugose-reticulate with thin rugae on the commissural face, less marked on 
the outer face, pale yellow. 

S. oregana is the only species of the section that is commonly an epiphyte and 
its long pendent branches are characteristic. Perhaps in relation to the length of 
the branches, the leaf-base is unusually long and the desiccated branches curl to 
form ringlets. The last two characters are sufficient for the identification of 
the occasional plants that grow on soil or rock. 

The specimen collected by Scouler (335, GH, NY) and frequently cited can 
not be accepted as coming from Observatory Inlet as stated on the label since this 
locality is so far north of the otherwise known range. 

In megaspore characters S. oregana shows a relationship to S. Underwoodii, 
which occasionally has pendent branches although it is not an epiphyte. 

Pendent from mossy trunks or branches of trees, particularly Acer macro- 
phyllum, or on shaded rocky banks, sea level to 200 m. 

Coastal "Washington to northern California. 

Representative specimens: 
United States, washingtc 

iorf ( 

(GH, MO, US); Thompson 5932 (GH, MO, NY, US), 624.1 (GH, MO), 9399 (GH, 
NY, US), 11425 (GH, MO, NY). Oregon: E. Hall 694 (F, GH, MO, NY, US); July 
12, 1882, T. Howell (F, GH, MO, NY, US); Fort [Port] Orford, 1855, Lieut. A. V. 


e? S. T. Parks 24119 (F, GH, NY, US). 

Howell 696 (GH, US) ; H. E. Parks 24274 

r , US) ; Summers 2209 
California: Eastwood 

31. Selaginella Underwoodii Hieron. in Engl. & Prantl, Nat. Pflanz. 1 4 :714. 
1901, based on S. rupestris var. Fendleri Underw. Figs. 39, 40. Map 41. 

Selaginella rupestris (L.) Spring var. Fendleri Underw. in Bull. Torr. Bot. Club 25:127. 

1898. (Lectotvoe bv Weath 


1944: Fendler 1 02 4 NY!. 

Selaginella Fendleri (Underw.) Hieron. in Hedwigia 39:303. 1900, not Baker, 1887. 




Paratypes: Ferriss in 1904 GH!; Good- 
745 GH!; Metcalfe 71 1 US! fragment 


Stems long or moderately long, forming open, spreading mats, rarely pendent 
and forming festoons; branches long and remote or moderately so, intricate, not 
fragmenting when dry. Upper and under leaves equal or subequal in length on 
the same portion of the stem; leaves subulate to linear to ligulate-long- triangular; 
base glabrous to pubescent, that of the upper leaves adnate to the stem for about 
Y 6 their length; margins ciliate, the cilia ascending toward the apex; apex of the 





[Vol. 42 


upper leaves herbaceous to slightly fleshy, slightly rounded to narrowly carinate, 
plane to abruptly beveled in profile; setae sometimes forming conspicuous tufts at 
the dry branch-tips, % as long to nearly as long as the blade, usually smooth, some- 
times scabrous, whitish, greenish-white to lutescent, rarely milk-white, usually 
translucent to rarely opaque. Sporophylls eciliate or with cilia dentiform and 
ascending toward the apex, the broadest usually 2 to rarely 3 times as broad as the 
leaves. Megaspores rugose-reticulate on the commissural face, more prominently 
marked, often with thin rugae, on the outer face, pale orange. 

On the basis of the material I have studied I am not able to recognize var. 
dolichotricha. It is a considerably less well-defined entity than others I have rec- 
ognized as varieties. S. Underwoodii from northern Arizona, northern New Mexico 
and northward is rather uniform in having short and well-differentiated setae and 
mostly short cilia on the leaves. However, to the south this extreme is also found 
with the phase that has long and poorly differentiated setae and long cilia. The 
three characters do not correlate as well as one would like in this area. This is a 
close parallel to S. mutica but in that species the differentiation has proceeded 
further and varieties are recognized. 

S. Underwoodii often grows with S. mutica and as mentioned under that species 
the S. Underwoodii in such mixed mats eventually disappears. 

Moist or shaded cliffs or rocky slopes, rocky alpine meadows, in crevices or on 
ledges or among rocks, usually on granitic or other igneous rocks, less often on 
sandstone or limestone, 800-4000 m., usually from 2000-3000 m. 

Representative specimens: 


United States. Wyoming: Payson 2503 (US). Colorado: C. F. Baker 2 (NY, US); 
C. S. Crandall 1054 (NY, US); Johnston 2424 (GH, US), 2425 (GH, NY, US); R. M. 
& A. F. Try on 5074 (ARIZ, B, BM, CU, F, FI, GH, K, MO, NY, P, PH, RM, UC, US, 
WS), 5076 (BM, DS, GH, MIN, MO, POM, US), 5078 (B, BM, F, GH, MICH, MO, 
NY, P, US), new Mexico: Arsene 15872 (F, MO), 17963 (F, US); 1874, Fendler 1024 
(B, F, GH, MO, NY, US) ; Mogollon Mountains, Mogollon Creek, Socorro Co., 8000 ft., 
July 20, 1903, Metcalfe 276 (GH, MO, NY, US); Metcalfe 711 (GH, MO, NY, US), 
991 (F, GH, MO, NY, US) ; Richards 8 Drouet 363 (F, GH, MO, NY) ; March 1, 1892, 
Wooton (NY). Oklahoma: Aug. 14, 1937, Wherry (US). Texas: Ferris & Duncan 3588 
(NY, US); Hinckley ion (GH, US), 1 1 56 (US). Arizona: Darrow tf Phillips 2781 
(GH, MO, US); March, 1904, Ferriss (GH) ; Goodding 5 (GH, US), 213 (GH, NY); 
Maguire et at. 11745 (GH, US) ; Phillips 2866 (GH, MO, US) ; Pultz & Phillips 2749 

32. Selaginella rupestris (L.) Spring, in Mart. Fl. Bras. 1 2 :118. 1840. 

Figs. 41, 42. Map 42. 

Lycopodium rupestre L. Sp. PI. 2:1101. 1753. (Lectotype: Kalm LINN!. Kamtchatcha, 

Steller LINN! is excluded as a type; it is S. sibirica). 
Stachygynandrum rupestre (L.) Beauv. Prod. Aethiog. 110. 1805. 
Selaginella Bourgeaui Hieron. in Hedwigia 39:295. 1900, as Bourgeauii. (Lectotype: 

Bourgeau on Aug. 15, 18 57 B!; Black Hills, Hayden MO! is an excellent match. Para- 

type: Bourgeau on Aug. 14, 1857, in part, B! fragment NY!). 
iginella rupestris f . Bourge 




Stems long to rather short, forming open, spreading mats; branches long and 
remote to moderately short and subapproximate, intricate, rarely only slightly so, 
not fragmenting when dry. Upper and under leaves equal or subequal, rarely the 
under slightly longer, on the same portion of the stem; leaves subulate to ligulate- 
long- triangular; base usually pubescent, rarely glabrous, that of the upper leaves 
adnate to the stem for % to ^4 their length; margins ciliate, the cilia spreading 
toward the apex; apex of the upper leaves herbaceous, rounded, gently to abruptly 
beveled in profile; setae forming conspicuous tufts at the dry branch- tips, nearly 
Vz to % as long as the blade, scabrous, usually strongly so, milk-white to tawny, 
opaque to rarely subopaque. Sporophylls with the cilia piliform and spreading 
toward the apex, rarely laxly ascending, the broadest about 4 times as broad as the 
leaves. Megaspores usually 1—2, less often 3 or 4 in a sporangium, rugose to rugose- 
reticulate, more finely marked on the outer face, bright orange. 

S. rupestris is the only species that is certainly apogamous. In the Appalachian 
mountain region as far north as southern Pennsylvania most of the material has 
four megaspores in a sporangium and microsporangia are present in the strobilus. 
Such plants are presumably sexual. In some plants the spore number in the 
sporangia of a strobilus may vary from 1 to 4. Those having 1-2 megaspores in 
the sporangia and a few microsporangia are less common. Throughout the rest of 
the range the most frequent type bears only megasporangia with either 1 or 2 
megaspores. This must be an obligate apomict and it is the only kind in a broad 
band on the northern periphery of the range. From Missouri to Michigan and 
Vermont there rarely occur specimens that bear strobili having a few micro- 
sporangia. The widespread apogamy may explain the uniformity of S. rupestris 
over its broad range. All other wide-ranging species are much more variable. The 
occasional occurrence of 1-2 megaspores in sporangia of S, densa and S. sibirica may 
also indicate apogamy. 

S. rupestris most closely resembles S. densa var. densa and it can best be sep- 
arated by its radially symmetrical leafy stem. In addition, in the area in Canada 
where the two grow together, all of the S. rupestris have one or two megaspores in 
a sporangium while S. densa has, with rare exception, four. When S. rupestris 
grows in nearly pure sand and other vegetation is sparse it will form circular, or 
in age, ring-shaped mats. 

Exposed or less often shaded cliffs, rocky bluffs, of acidic igneous or sedimentary 
rocks, gravel or sandy soil, up to 1900 m. 

Central and eastern United States to northeastern Alberta, Quebec and Green- 

Representative specimens: 

Greenland: July 13, 1946, T. W. Bocher (GH). 

Canada, nova scotia: Fernald & Long 23098 (GH, US). Quebec: Victorjn 8421 
(MO, US) ; Victorin et al 4201 (MO, US). Ontario: C. £. tf G. K. ]ennings 7362 (GH, 
US), 7389 (GH, US); July 20, 1899, Vmbach (MIN, NY, US); Van Eseltine 501 
(GH, US). Manitoba: Fort Ellice, Aug. 15, 1857, Bourgeau (B), Aug. 14, 1857 (B, 
NY); Macoun 137 (BM); Macaun tf Herrht 70372 (B, F, GH, NY). Saskatchewan: 

[Vol. 42 


Roup 6305 (GH, NY), 6929 (GH, NY), alberta: Raup 6 Abbe 4451 (GH, NY), 

4608 (GH, NY). 

United States, maine: Hodgdon (Pi. Exsicc. Gray. 611) (F, GH, MO, NY, US). 

Vermont: Sept. 11, 1920, Dutton (F, GH, MO) ; Eggleston 2199 
necticut: 1859. D. C. Eaton 


NY, US), 32 (F,GH,MO,NY,US). Pennsylvania: Heller tf Halbach 

US). Virginia: Steele & Steele 166 (GH,MO,NY,US). north Carolina: Biltmore Herb. 

MO, NY). 

lin a: Mackenzie 2997 (MO, US) ; ( 
98 (US) ; Correll 6615 (GH, MO, 1 
?40? (F, MO). Michigan: R. W 

US); R. M. tf P. F. Tryon 4695 (GH, MO). Indiana: June 16, 1900, Umbach (GH, 


(F, GH, MO, NY, US), wis- 
), P, US). Illinois: A, Chase 

consin: K. M. & A. t. Tryon 5005 (b, M, GH, K, MO, P, US). 

/5<?^ (F, US); Evers 32650 (MO). Minnesota: Bergman 3156 (GH, NY, US); Moyfe 
483 (F, GH, MO, US). Missouri: Bush 4744 (GH, MO, NY, US), 5210 (GH, NY, US). 

Arkansas: E. /. Palmer 35561 (F, GH, MO, NY), south dakota: Murdoch 4304 (F, 
GH, US). Nebraska: Kiener 1 1 323 (US), 23695 (MO). Kansas: 1886, E. N. Plank 14 
(GH). Oklahoma: Bush 835 (GH, MO, NY). 

33. Selaglnella densa Rydb. in Mem. N. Y. Bot. Gard. 1:7. Feb. 15, 1900. 
(Holotype: Havard NY! marked as type by Rydb., fragment US!. Paratypes: 
Newberry NY!; Williams 534; Tweedy 172; Missoula, Mont. 1898, Williams 
& Griffith; Silver Bow Co., Mont., Mrs. Jennie Moore; the last four collections 
probably at MONT. Watson in 1880 GH! is excluded as type material; it is 

S. Wallacei) . 

Map 43. 

Stems short, forming flat cushion mats; branches short, approximate, discrete, 
not fragmenting when dry. Under leaves definitely longer than the upper on the 
same portion of the stem, or sometimes equal to the upper on the assurgent branch- 
tips; leaves ligulate to ligulate-lanceolate to ligulate-long- triangular; base glabrous 
to pubescent, that of the upper leaves adnate to the stem for % to % their length; 
margins eciliate or with the cilia ascending to spreading toward the apex; apex of 
the upper leaves herbaceous to fleshy, slightly to broadly rounded, plane to truncate 
in profile; setae forming conspicuous tufts at the dry branch- tips, % to % as long 
as the blade, smooth to scabrous, milk-white and opaque to tawny or whitish and 
subopaque to lutescent and translucent. Sporophylls eciliate or with the cilia 
dentiform to piliform and ascending, rarely laxly ascending, toward the apex, the 
broadest 3 to 4 times as broad as the leaves; seta base often strongly broadened and 
flattened. Megaspores prominently and coarsely rugose-reticulate to slightly 
rugose, most prominently marked in the equatorial region, pale to bright orange. 

S. densa is one of the most complex species of the section. The morphological 
extremes of the three varieties are amply distinct, but from Montana south to 
Colorado all occur and with an increasing number of intermediates. In Colorado 
nearly all possible intermediate conditions between the three may be found and 
over a third of the collections I have seen from there are such intermediates. To 
the south, in the mountains of Arizona, New Mexico and Texas, there is less 
variation. Var. Standleyi and var. densa are lacking and there are no intermediates 
with var. Standleyi. The material there is either typical var. scopulorum or inter- 
mediates of it varying toward var. densa. 



It is not possible to withhold recognition from the three extremes since they do 
have distinctive characters and distribution in the northern part of the range of 
the species. At the same time, it must be admitted that identification, particularly 
of the material from Colorado, is sometimes rather arbitrary. The intermediates 
show all transitions between the two extremes in respect to the varietal characters. 
Some specimens will be intermediate in one character, others in two or three. Some 
will have some of the leaves and sporophylls intermediate, others typical of one 
variety; or some leaves and sporophylls will be typical of one variety, others of 
them typical of another variety. 

The intermediate specimens have, for the purposes of discussion, been divided 
into two kinds. Those that are more or less halfway between the typical state of 
two varieties are called intermediate. Those that depart from the typical state of 
a variety to a lesser degree are mentioned as specimens showing a tendency toward 
another variety. Such specimens can be identified with one variety quite satis- 
factorily but are not entirely typical of it. 

Of the three varieties, var. densa most closely resembles S. rupestris and the 
differences are discussed under that species. Rarely S. densa may have some mega- 
sporangia with two megaspores but this has not been found with sufficient reg- 
ularity within a strobilus to indicate functional apogamy as in S. rupestris. 

Southwestern Manitoba to southern Alaska, south to Texas, Arizona and 
northern California. 


a. Apices of the leaves and sporophylls plane to abruptly beveled in profile; 
setae of the leaves white or whitish, or lutescent only at the base, opaque 
to subopaque. b. 

b. Sporophylls eciliate toward the apex 33a. var. scopulorum, p. 67 

b. Sporophylls dentiform to piliform-ciliate to the apex _. 3 3b. var. densa, p. 68 

a. Apices of the leaves and sporophylls predominantly or all truncate in pro- 
file; setae of the leaves whitish- to greenish- to entirely lutescent, trans- 
lucent 33 c. var. Stand ley i, p. 71 

/33a. Selaginella densa var. scopulorum (Maxon) Tryon, comb. nov. 

Fig. 43. Map 44. 

Lycopodrum bryoides Nutt. ex Baker, Handb. Fern Allies, 35. 1887, in synon. Placed here 
on the basis of the specimen so labeled at Kew! (fragment NY!) which agrees with 

Baker's description. 
Selaginella scopulorum Maxon, in Amer. Fern Jour. 11:36. 1921. (Holotype: Standley 

15732 US!. Paratypes: Ferris & Dutbie 941 US!; Flett 3092 US!; Goodding 483 US!; 

Heacock 235 US!; Mearns 4274 US!; Merrill & Wilcox 1 2 18 US!; Shaw 398 USl, 902 

US!, 1060 US!; Standley 15318a US!, 15398 US!, 16216 US!, 16255 \5S\,l6288 US!, 

16378 US!, 17055 US!, 17164a US!, 17979 US!, 18050 US!, 18185 US!; Suksdorf 

8834 US!; Ulke on Aug. 25, 1917 US!; Umbacb 856 US!). 
Selaginella columbiana A. A. Eaton ex Maxon, in Amer. Fern Jour. 11:37. 1921, in iynon. 

(Evidently based on Shaw 398 and Heacock 235). 

[Vol. 42 


Apex of leaves plane to abruptly beveled in profile; setae white, or lutescent 
only at the base, opaque, rarely subtranslucent. Sporophylls eciliate toward the 
apex; apex plane to gently beveled in profile. 

Var. scopulorum is apparently the least specialized of the varieties of S. densa. 


frequent but intermediates with var. Standleyi are rare. 


densa. New Mexico, Arsenetf Benedict 16370 (US) ; Arizona, 

2Q00 (GH); Colorado, Underwood & Selby 133 (NY); Utah, Hermann 5063 


Examples of specimens of var. scopulorum with a tendency toward var. densa 
are: Colorado, Knowlton 82 (US), McKelvey 4695 (GH) ; Utah, Harrison 6 
Larsen 7891 (MO) ; Arizona, Phillips & Reynolds 2900 (US). 

Colorado, Kydberg & Vreeland 6588 (NY) represents var. scopulorum with 

a tendency toward var. Standleyi. 



are mentioned under var. densa. 

Usually in rocky alpine tundra, also cliffs, talus slopes, on ledges, among 
boulders or in thin soil over rocks, on igneous or sedimentary rocks, 700-4660 m. 

Alberta to British Columbia, south to Texas, Arizona and northern California. 

Representative specimens: 

Canada, alberta: Scamman 2790 (GH); Rosendahl 1074 (US). British Columbia: 
Heacock 235 (BM, GH, MO, NY, US); Hitchcock tf Martin 7435 (GH, MO, NY); 
Shaw 398 (BM, GH, MO, NY, US), 902 (BM, GH, NY, US), 1060 (BM, GH, MO, NY, 

United States. Montana: Standley 15318a (US), 15598 (US), vicinity of Cracker 
Lake, Glacier National Park, 1740-1920 m., July 15, 1919, Standley 1 57 32 (US), Stand- 
ley 16216 (US), 16255 (US), 16288 (US), 16378 (US), 17055 (US), 17164a (US), 
17979 (US), 18050 (US), 18185 (US); Aug. 25, 1917, Ulke (US); Vmbach 856 (F, 
MIN, NY, US). Wyoming: Goodd'tng 483 (F, GH, MO, NY, US); Mearns 4274 (US); 
Merrill & Wilcox 1218 (GH, NY, US). Colorado: Cox 290 (F, MO); Murdoch 4787 

(F, US), utah: E. B. tf L. B. Payson 4037 (GH, MO). Texas: Moore & Steyermark 
3245 (US), new Mexico: Metcalfe 1 172 (BM, F, GH, MIN, MO, NY, US). Arizona: 

1913, Ferriss (US), idaho: Epling 7011 (F, MO, US). Washington: Flett 

Suksdorf 81 
Holmgren 2 

3036 (US). 

S (F, MO, US). Washington: Flett 3092 (US); 
Oregon: Ferris & Duthie 941 (US) ; Maguire & 
fornia: J. T. Howell I1S2I (US) : L. C. Wheeler 

33b. Selaginella densa var. densa. Figs. 44, 45, 47. Map 45. 

Selaginella longipila Hieron. in Hcdwigia 39:291. Dec. 28, 1900. (Lectotype: Bourgeau 
in 1858 (Herb. Kew Exsicc. no. 1239, wrongly labeled "Himalaya") B!. Paratype: 
"Herb. Griffith, Bhotan" B! fragment NY!, collector and locality unknown). 

Selaginella rupestris (L.) Spring f. longipila A. Br. ex Hieron. in Hedwigia 39:291. 1900, 
in synon. 

Selaginella Engelmannii Hieron. in Hedwigia 39:294. Dec. 28, 1900. (Holotype: Engel- 
mann B!). 








[Vol. 42 

Selaginella Haydenii Hieron. in Hedwigia 39:296, Dec. 28, 1900, as Haydeni. (Lecto- 

type: Hay den B!. Paratype: Lyall B! fragment NY!). 
Selaginella rupestris var. densa (Rydb.) Clute, Fern Allies, 142. 1905. 
Selaginella rupestris f. Engelmannii (Hieron.) Clute, in Fern Bull. 16:52. 1908. 
Selaginella rupestris f. Haydenii (Hieron.) Clute, in Fern Bull. 16:52. 1908. 

Apex of leaves plane to abruptly beveled in profile; setae milk-white and 
opaque to whitish and subopaque. Sporophylls with the cilia dentiform to piliform 
and ascending, rarely laxly ascending, toward the apex; apex plane to abruptly 
beveled in profile. 

misapplication of S. longipila 


i been discussed under S. indica. The type specimens are 
actually S. densa var. densa but fortunately Rydberg's well-known name has a 

few months priority. 

under var 

pulorum are mentioned under that variety and 


26, 1896, Holm (MO) 

July 6, 1937, Wherry 




An example of var. densa with a tendency toward var. Standleyi is: Empire, 
Colorado, Aug. 27, 1874, Engelmann (B, MO, US). 

The following specimens are rather intermediate between all three varieties: 
Wyoming, Hermann 4665 (MO) ; Colorado, Ewan 12699 (Ewan) ; New Mexico, 
Arsene & Benedict 18056 (F) . 

Prairies, alpine meadows or dry rocky places, acidic rocks or sandy soil, 1100- 

4000 m. 

Southeastern Manitoba to British Columbia, south to New Mexico and Ar 
Representative specimens: 


Canada Manitoba: Macoun ef Her riot 70373 (F,GH,NY). Saskatchewan: Boivin 
fif GtUett 8686 (MO); Fort Carlton, March 29, 1858, Bourgeau (Herb. Kew Exsicc. no. 
1239) (B, NY, P, US, YU) ; Cowles 43 (F, MO) ; Gillett 6022 (MO), alberta: Breitung 
5553 (MO); S. Brown 128 (GH, MO, US); Make tf Watson 1229 (GH). British 
Columbia: Calder & Savile 7677 (MO), 9221 (MO). 

United States. "Oregon", 49 N. Lat., 1858-59, Lyall (B, NY), north Dakota: 
Aug. 12, 1908, Lunell (GH, MIN, US), Sept. 10, 1908 (NY, US), south Dakota: Black 
Hills, 1853-54, F. V. Hayden (B, MO, NY, US); Hay-ward 315 (F, NY). Montana: 
Little Rocky Mountains, Sept. 1889, Havard (NY, US); Newberry (NY)- Rydberg & 
Bessey 3517 (F, GH, MIN, NY, US); R. S. Williams 534 (US). Wyoming: A. Nelson 

6590 (NY) ; L. tf R. Williams 2126 




^33c. Selaginella densa var. Standleyi (Maxon) Tryon, comb. nov. 

Fig. 46. Map 46. 

Selaginella Standleyi Maxon, in Smiths. Misc. Coll. 72 5 :9. 1920. (Holotype: Standley 
17228 US!. Paratypes: Brown 95 US!; Standley 15363 US, 16970 US!, 17483 US!, 
18 1 36 US!; Ulke on Aug. 25, 1917 US!). 

Apex of leaves predominantly or entirely truncate in profile; setae whitish- to 
greenish-lutescent to lutescent, translucent. Sporophylls eciliate or with the cilia 
dentiform or piliform and ascending toward the apex; apex truncate, rarely 
abruptly beveled in profile. 

Intermediates of var. Standleyi and var. densa and material of var. densa with 
a tendency toward var. Standleyi are mentioned under var. densa. Specimens of 
var. scopulorum with a tendency toward var. Standleyi are mentioned under var. 

Specimens of var. Standleyi with a tendency toward var. densa are such as: 

ndley 16970 (US), Van Schaack 2786 l / 2 (MO); Colorado, Johnston 



J. H. Schaffner (BM) ; Colorado, Underwood 



Southern Alaska to Colorado. 
Representative specimens: 

Alaska: A. & A. Krause 162 (B). 


Holway & Butters (MIN, US) ; Raup & Abbe (GH, NY). 

United States. Montana: Barkley 1725 (GH, MO, US); Standley 15363 (US), 
vicinity of Sexton Glacier, Glacier National Park, Aug. 7, 1919, Standley 17228 (US), 
Standley 17483 (US), 18136 (US); Aug. 25, 1917, Ulke (US). Colorado: Johnston 


Hedwieia 39:290. 1900. 

Fig. 48. Map 47. 

Selaginella rupestris (L.) Spring f. sibirica Milde, Fil. Europ. Atlant. 262. 1867. (Lecto- 

type: Unalaska, Chamisso. Paratypes: Ajan, Tilling; Dahuria ad flumen Ingoda, Pallas). 

Selaginella rupestris f. amurensis Milde, Fil. Europ. Atlant. 262. 1867. (Holotype: Amur, 


_:. manchu 
Khalkyli, Maack) . 

1867. (Holotype: 

- j w j ^ - ^— — w w— — w -mm m w 

Selaginella Schmidtii Hieron. in Hedwigia 39:292. 1900. (Lectotype: Schmidt B! frag- 
ment NY!. Paratype: Chamisso B! fragment NY!). 

Selaginella Schmidtii var. Krauseorum Hieron. in Hedwigia 39:293. 1900. (Lectotype: 
A. & A. Krause 53 B! fragment NY!. A. ti A. Krause 162 B! is excluded as a type- 
it is S. densa var. Standleyi) . 

Selaginella rupestris i. Schmidtii (Hieron.) Clute, in Fern Bull. 16:52. 1908. 

Stems long to moderately long, forming open, spreading mats; branches long 
and remote to moderately short and subapproximate, intricate, not fragmenting 

* m . _- . « » • 1 . 1 - 1 _ _ J 

when dry. 

bequal in length, or the under 

[Vol. 42 


slightly longer, on the same portion of the stem; leaves linear to ligulate-long- 
triangular; base usually glabrous, less often pubescent, that of the upper leaves 
adnate to the stem for % their length; margins ciliate, the cilia spreading to laxly 
ascending toward the apex; apex of the upper leaves fleshy, broadly rounded to 
carinate, subtruncate to usually truncate in profile; setae forming conspicuous 
tufts at the dry branch-tips, % to % as long as the blade, scabrous, milk-white 
to white to tawny, opaque to translucent. Sporophylls eciliate or with the cilia 
piliform to dentiform and ascending toward the apex, the broadest 3 to usually 4 
times as broad as the leaves; seta base usually strongly broadened and flattened. 

Megaspores rugose to rugose-reticulate, pale yellow to pale orange. 

The Alaskan and Yukon material is relatively uniform while that from Asia 
is less so. In addition to the typical form with short, milk-white setae a phase 
occurs in Asia with longer and tawny setae, and also occasional specimens bear 
strobili having 1 to 2 megaspores in a sporangium. This is quite parallel with 
S. densa var. densa, the reduced number of megaspores being so rare that it is not 
possible to be certain of apogamy. 

S. sibirica is most similar to S. densa var. Standleyi from which it may be sep- 
arated by the white to tawny rather than lutescent setae and the intricate rather 
than discrete branches. It is separated from S. shakotanensis, which also grows in 
Japan, by the leaf-bases that place the two in different series. The seta length 
likewise distinguishes these species — in S. sibirica they are %— % as long as the 
blade while in S. shakotanensis they are I/5 as long. 

The only sporelings observed in this study were seen in the soil of a mat of 
S. sibirica collected by Calder & Billard (2995 MO). These are discussed in some 
detail in the introduction. 


Representative specimens: 



4454 (MO), 460 

Alaska: unalaska, Chamisso (B, NY) ; Flett 1529 (NY, US) ; A. E. & R. T. Porsild 
690 (GH, MO, US); Scamman 198 1 (GH, MO). 

Behring Straits: Emma Harbour, Sept. 21, 1881, A. # A. Krause 53 (B, NY); 
C. Wright (GH, NY, US, YU) . 

Union of Soviet Socialist Republics, kamtchatka: Eyerdam (F) ; Novogablenov 
541 (US), sachalin: Fr. Schmidt (B, GH, NY, US), trans Baikal: Ingoda river 
(Dahuria), Fischer (GH, P). 

1 100 (P), 7279 (P) 9 S4g6 (MO, P), 9687 (MO, P), 13903 

35. Selaginella Watsonii Un 
Watsoni. (Holotvoe: Watson 

1898, as 


Hansen 879 NY!; Coville & Funs ton 2071 NY!; Brewer 2103 NY!). 

Figs. 49-51. Map 


subapproximate to moderately long and remote, intricate, not readily fragmenting 


Upper and under leaves equal in 




length, or the under slightly longer, on the same portion of the stem; leaves ligulate 
to broadly ligulate to ligulate-long- triangular; base glabrous to slightly pubescent, 
that of the upper leaves adnate to the stem for % to % their length; margins 
eciliate, rarely with ascending cilia toward the apex; apex of the upper leaves fleshy, 
narrowly to broadly carinate, subtruncate to truncate in profile; setae not or hardly 
forming conspicuous tufts at the dry branch-tips, */$ to %, rarely x fe y as long as 
the blade, usually smooth, sometimes slightly scabrous, greenish-white to greenish- 
lutescent to whitish-lutescent, translucent to subopaque. Sporophylls eciliate 
toward the apex, the broadest about 2 times as broad as the leaves. Megaspores 
rugose to rugose-reticulate, pale orange. 

S. Watsonii is characterized by its fleshy leaves which are usually eciliate and 
short-setate. In habit it forms dense rounded mats with intricate branches. The 
fragile nature of the stem of the next three species is poorly developed in S- 
Watsonii. The stems can be easily broken by hand but specimens do not fragment 
in packets or with ordinary handling. It seems probable that the next three species 
have been derived from S. Watsonii or a common ancestor. 

Exposed or shaded cliffs, talus slopes or rocky alpine meadows, in crevices or 
on boulders or among rocks, usually related to igneous rocks, rarely to limestone, 
1800-4300 m.. usuallv 3000- 

3700 m. 





[Vol. 42 

Representative specimens: 

United States. Montana: T. D. Howe 66 (US), utah: G. /. Goodman (GH, MO, 
NY, US) ; M. £. Jones 1246 (NY, US) ; Maguire tf Richards 13166 (GH, MO) ; E. B. (j 
L. B. Payson 4917 (GH, MO, US), 5074 (GH, MO, NY, US) ; Rydberg 8 Carlton 6566 
(GH, NY, US); Cottonwood Canon, 9500 ft., July, 1869, S. Watson 1370 (NY, YU). 
Nevada: Heller 11054 (F, GH, MO, NY, US); Maguire 2IOQ0 (GH, NY, US); Sept., 
1868, S. Watson 1370 (NY, US). Oregon: July 1, 1931, Wherry (US). California: 
Brewer 2103 (NY, YU); Coville tf Funston 2071 (NY, US); Sept. 5, 1923, C. C. Hall 
(F, MO, NY, US) ; Hansen 879 (BM, MO, NY, US) ; Heller 7167 (GH, MO, NY, US) ; 
R. M. ef A. F. Tryon 5060 (ARIZ, B, BM, CU, DS, F, FI, GH, K, MICH, MIN, MO, 
NY, P, PH, POM, RM, UC, US, WS) . 

36. Selaginella utahensis Flowers, in Amer. Fern Jour. 39:83. 1949. (Holo- 
type: Cottam 5644 UT fragment US!. Paratype: Cottam 8817 UT). 

Fig. 52. Map 49. 

Stems moderately short, forming rounded to flat cushion mats; branches mod- 
erately short to moderately long, approximate to nearly remote, intricate, readily 
fragmenting when dry. Upper and under leaves equal or subequal in length on the 
same portion of the stem; leaves subulate to ligulate-lanceolate to ligulate-long- 
triangular; base usually glabrous, sometimes pubescent, that of the upper leaves 
adnate to the stem for l / 4 their length; margins eciliate to ascending-ciliate toward 
the apex; apex of the upper leaves slightly fleshy, broadly carinate, gently to rather 
abruptly beveled in profile; setae not forming conspicuous tufts at the dry branch- 
tips, less than % as long as the blade to essentially absent, smooth, whitish to 
greenish- to lutescent-whitish, subopaque. Sporophylls eciliate or with the cilia 
dentiform and strongly ascending toward the apex, the broadest 2 to 3 times as 
broad as the leaves. Megaspores slightly rugose-reticulate to slightly rugose, 

S. utahensis may be separated from the next two species, that also share the 
character of readily fragmenting stems, by its leaves which are muticous or have 
short, smooth setae. The anatomical basis of the fragmenting stems has not been 
investigated. In mounted material the parts will be held together if glue or soil 
holds the roots and rhizophores firmly but in packets or in unmounted material 
ordinary handling of the sheet will soon reduce the stems to small pieces. 

Ledges and crevices of sandstone cliffs, 1500-2300 m. 

Southern Nevada and southwestern Utah. 

Specimens examined: 

United States, utah: Lady Mountain, Zion National Park, 6,500 ft., April 5, 1931, 
W. P. Cottam 5644 (MO, US) ; Degener & Peiler 16982 (NY) ; Eastwood ef Howell II 59 
(US); Flowers 3249 (MO). Nevada: June 28, 1930, E. Jaeger (US). 

37. Selaginella leucobryoides Maxon, in Smiths. Misc. Coll. 72 5 :8. 1920. 
(Holotype: Munz & Harwood 3789 US!. Para types: Coville & Funston 628 
US!; Munz, Johnston d Harwood 4226 US!). Fig. 53. Map 50. 

Stems short, forming flat or rounded cushion mats; branches short, approxi- 
mate, intricate or discrete, readily fragmenting when dry. Upper and under leaves 



equal in length, or the under slightly longer, on the same portion of the stem; 
leaves linear to ligulate-ovate; base glabrous to pubescent, that of the upper leaves 
adnate to the stem for l / 4 or less their length; margins eciliate to ascending-cilia te 
toward the apex; apex of the upper leaves herbaceous, slightly to broadly rounded, 
gently to abruptly beveled in profile; setae not forming conspicuous tufts at the 
dry branch-tips, % or less as long as the blade, scabrous to slightly scabrous, milk- 
white and opaque, rarely white and translucent. Sporophylls eciliate or with the 
cilia dentiform and ascending toward the apex, the broadest about 3 times as broad 
as the leaves. Megaspores slightly rugose to rugose-reticulate, to nearly smooth on 
the outer face, pale orange. 

The short, white, scabrous and usually opaque setae are characteristic of S. 
leucobryoides. It shares the peculiar character of fragile stems with S. utahensis and 

S. asprella. 

Rocky slopes, in crevices or among rocks, 900-2300 m. 
Southeastern California. 
Specimens examined: 

United States. California: Coville & Funs ton 628 (NY, US); Coville & Oilman 
III (US); /. T. Howell 3989 (F, GH, MO, US); Bonanza Mine, Providence Mountains, 
2800 ft., March 30, 1920, P. A. Munz & R. D. Harwood 3789 (F, GH, NY, US) ; Munz, 
Johnston tf Harwood 4226 (US). 

38. Selaginella asprella Maxon, in Smiths. Misc. Coll. 72 5 :6. 1920. (Holo- 

type: Johnston 18 15 US!. Paratypes: Johnston 1595 US!, 1807 US!). 

Fig. 54. Map 51. 

Stems short to moderately short, forming rounded or flat cushion mats; 
branches moderately short and remote to short and subapproximate, intricate, rather 
readily fragmenting when dry. Upper and under leaves essentially equal in length 
on the same portion of the stem; leaves linear-lanceolate to lanceolate-ovate to 
lanceolate-long-triangular; base pubescent, rarely glabrous, that of the upper leaves 
adnate to the stem for about % their length; margins eciliate to ascending-ciliate 
toward the apex; apex of the upper leaves herbaceous, broadly rounded to carinate, 
nearly plane to truncate in profile; setae forming conspicuous tuhs at the dry 
branch-tips, % to over l / 2 as long as the blade, scabrous, white and translucent to 
rarely milk-white and opaque. Sporophylls eciliate or with the cilia piliform to 
dentiform and strongly ascending toward the apex, the broadest about 3 times as 
broad as the leaves. Megaspores prominently and coarsely rugose-reticulate, most 
prominently marked in the equatorial region, pale yellow to pale orange. 

S. asprella is characterized by its long, scabrous, white, usually translucent 
setae. Of the three species with readily fragmenting stems, it is the furthest re- 
moved from the presumed ancestor, S. Watsonii. 

Open areas in rocky soil or in crevices at the base of boulders, igneous rock, 


Southern California. 
Specimens examined: 

[Vol. 42 

United States. California: /. T. Howell 5031 (US); Jaeger 276 (US); /. M. 
Johnston 1595 (MO, US), 1807 (MO, US), West end of Ontario Peak, San Antonio 
Mountains, 6000 ft., March 25, 1918, /. M. Johnston 1815 (MO, US) ; Munz 7612 (NY), 
9683 (US), 17165 (GH); Sept. 14, 1921, Saunders (US); April, 1906, Streeter (NY, 



Series Eremophilae Tryon, ser. nov. 

Rhizomata et stolones et gemmae nulla. Caules prostrati. Apices ramorum 
involuti statu inerte. Caules frondosi valde dorsiventrales. Folia supera base 
abrupte adnata caule distincta colore, folia infera base valde decurrente. Typus: 
Selaginella eremophila Maxon. 

Plants terrestrial; rhizomes, stolons and basal buds absent. Stems prostrate or 
assurgent at the branch tips, with rhizophores produced generally throughout; 
stems long to rather short, forming open or usually rather compact, rounded or 
usually flat mats. Branches moderately long to short, remote to subapproximate, 
usually intricate, branch tips involute in the dormant state. Leafy stems strongly 
dorsiventral, the under leaves appressed, the upper erect, the under longer, usually 
thinner and different in shape from the upper; zone of green leaves much longer 
on the upper side of the stem, very short to absent on the under side. Upper leaves 
with the base abruptly adnate, distinct from the stem in color, under leaves with 




is increasing specialization in the dorsiventral habit. S. peruviana is stron 

ventral in position, length and shape of the upper and under leaves; in S. 

they are also different in texture. In S. eremophila the setae are highly 1 m 

and the leaves of S. Parishii are muticous. Finally, in S. Landii, the under leaves 

are unusually long in relation to the upper and they are setate while the upper 

leaves are muticous. The species are all quite distinct with the exception of the 



a. All leaves setate (sometimes only in the bud) ; under leaves less than twice 
as long as the upper, b. 

b. Setae mostly or entirely persistent, stout, straight, c. 

c. Under leaves subulate, acuminate, broadest at or very near the base, 
rarely linear-lanceolate, not appreciably thinner than the herbaceous 
upper leaves, setae 0.5-1.0 mm. long. Oklahoma to New Mexico; 

Mexico; Peru to Argentina 39. S. peruviana, p. 77 

c. Under leaves lanceolate to lanceolate-ovate, broadest above the base, 
papyraceous, the upper fleshy, setae 0.1-0.3 mm. long. Texas to 

Arizona; Baja California 40. S. arizonica, p. 78 

b. Setae mostly early- deciduous (sometimes present only in the bud) , deli- 





a. Upper leaves, or all leaves muticous. d. 

d. All leaves muticous; under leaves about twice, or less, as long as the 
upper; upper leaves lanceolate to lanceolate-triangular... 42. S. ?amhn ? . 80 

d. Upper leaves muticous; under leaves predominantly short-setate, the 
setae persistent; under leaves over twice as long as the upper; upper 
leaves long-deltoid or with parallel sides toward the base 43. S. Landu, p. 81 

39. Selaginella peruviana (Milde) Hieron. in Hedwigia 39:307. 1900 

Figs. 55, 56. Map 52. 

Lycopodinm ciliatum Ruiz ex Spring, in Nouv. Mem. Acad. Roy Belg. 24 (Monog. Fan, 
Lycopod.):55. 1850, in synon. (evidently based on R«,z p^B). 

SrlaJnelU rupestris (L.) Spring f. Peruvian, i Milde, Fd. Eu~P- Atl^t « 3 1867 . £Holo 
type: Ruiz 9 8 B!; the lower right-hand specimen of W tikes Exped. J, Peru, Gttl 

Sel^neUatpestris var. mexicana Milde, Fil. Europ. Atlant. 263. 1867. (Aschenbom 

65 B! is so annotated by Milde or A. Br.). .-,,-. T ,t, r u <>2-417 1896 

Stla^lla rupees va, peruvun. (Mildo, H.eron ■ ,*« . Bo, £££f '£,!£„ 
Srl.gwrHa Aichenbornii Hieron. in Hedwig.a MiJOS. 1? 00 ' l "'™ £ A4 ,.„„ ,,, B!; 

05 B!; PaW 554 GH! and US- (62778) ,s '^""''Vv Ja^/b Tolut, r™ 

gW «1 £U B, H* i. ..« B, Meaicc , C, y, SrW* B, Toluu, . . 

Chrismar B. Vancouver Island, /. ti. nenry d snouiu 


Bang III B!. 

Dombeyana Hieron. in Hedwigia_39:308. 1900- (Lectjype: 

Ba*g 171 B!. Paratypes: M<.»do» <3 4 *»5 ^"-^ *£ -• — ; ' . , lg74 

1892, O. Kuntze B; Argentina, Hi,ra«^«s tf Lorentz 162 B, Argentma, Jan. 

Lorentz & Hieronymus V) . _ . M , w1 ,010 as sheldoni. (Holo- 

Srb*f«tfe SfeMmti Maxon, in Proc. Biol. Soc. Wash. 31:171. 1918 a . U*ao ^ 

type: Sheldon 2 33 US!. Paratypes: F. C Grm,, on June 10, 1918 US., M 7 
1918 US!; Jenny 343 US!; HW US!; Wooton US!). 

Upper leaves herbaceous, or fleshy only at the base, linear, acuminate to U» **- 
long-triangular; under leaves herbaceous, subulate, acummate broadest at or vej 
near the bL, to rarely linear-lanceolate, a little longer than the upper on th same 
portion of the stem; leaves setate in the bud, setae percent (part may be ^abraded) , 
stout, straight, those of the under leaves 0.5-1.0 mm. long, rarely some 0.3 «£ 
long. Megaspores rather coarsely rugose-reticulate on the commissural face, less 
prominently marked on the outer face, yellow to bright orange 

r J -«..i__ • ~~A tr\r rhi<; reason it 

5. peruviana is the least specialize «-- — „ rttT ,;„^ (<stan- 

i, consLred as the basic type. Some (out ^^y^^lZZvS)^ 
for, e, al. „S MO, S te y e r,nark s»f6 F, An*» ^ * _^ ^ US) 


(0. 2 -0,' mm .) set,. T. ^. t:l?C'^^ 
under leayes, the long setae and the sumlar texture of PP 

The localities taken from the literature (map 52) are trom 

15:253. 1939. J in crevices of cliffs, eaposed or under light 

Bluffs, rocky slopes or on ledges and m crevices of , F ^ ^ 

shade, igneous rocks or sandstone less of ten ■ .sandy or « ay 


[Vol. 42 


Oklahoma and New Mexico to Puebla; Peru to Argentina. 
Representative specimens: 

United States. Oklahoma: Goodman 2337 (GH, MO, NY, US) ; May 4, 1918, F. C. 
Greene (US), June 10, 1918 (US); Quanah Mountain, Indian Territory, July 28, 1891, 
Sheldon 233 (US), texas: July, 1883, Havard (US); Jermy 343 (MO, US); Lindheimer 
77 (GH, MO, US) ; C. H. Mueller 8256 (GH, MO, NY, US) ; R. M. 6 A. F. Tryon 5032 
(BM, CU, DS, GH, MIN, MO, PH, POM, US, WS) ; March 7, 1918, Wooton (US). 
new Mexico: Arsene 18050 (F, US), 18613 (F, US); Arsene & Benedict 16838 (F, US). 

Mexico. Aschenborn 65 (B). coahuila: Stanford et al. 118 (GH, MO, NY, US). 
chihuahua: E. Palmer 38 in 1908 (GH, US). 


'■ffner 933 (GH, YU). distrito federal: Schaff 

608 (US) ; Dombey 

3 1 41 


Bolivia: Vicinity of La Paz, 10,000 ft., 1889, Bang m (B, F, GH, K, MO, P, US) ; 

Buchtien 4299 in 1919 (F, MO, US); Eyerdam 24798 (F); Mandon 84 (B), 1531 (B). 

Argentina: Burkart 10112 (MO); Hieronymus & Lorentz 162 (P) ; Lossen 426 
(F, MO). 

40. Selaginella arizonica Maxon, in Smiths. Misc. Coll. 72 5 :5. 1920. (Holo- 
type: Shreve US!. Paratypes: 7 bomber 315 US!; Cook US!; Parish 8513 US!; 
P. F. Mohr US!; Tourney US!; Good ding 722 US!; Bailey in 1913 US!, on 

Nov. 9, 1913 US!). 

Figs. 57, 58. Map 53. 

Upper leaves fleshy, lanceolate to linear-lanceolate; under leaves papyraceous, 
lanceolate to lanceolate-ovate, broadest above the base, about as long as to slightly 
longer than the upper on the same portion of the stem; leaves setate in the bud, 
setae predominantly (or many) persistent, stout, straight, those of the under 
leaves 0.1-0.3 mm. long. Megaspores rather finely rugose-reticulate on the com- 
missural face, less prominently marked on the outer face, pale orange. 



arizonica in 
vo sometimes 


separable. Some strobili of A. & R. A. Nelson 1158 MO have vegetative growth 
beyond the tip, a condition not uncommon in S. arenicola ssp. Riddellii. 

Exposed rocky places, in crevices or on ledges, or in gravel, usually on igneous 
rock, rarely on limestone, 600-2000 m. 

Southwestern Texas, Arizona and adjacent Sonora, Baja California. 

Representative specimens : 

United States, texas: Cory 6878 (GH), 40232 

on 5048 

PH, POM, RM). Arizona: Nov. 9, 1913, V. Bailey (US), 1913, V. Bailey (US); 1913, 


11217a (GH, NY, US) ; Parish 8513 (US) ; Phillips 2930 (GH, MO, US) ; Foot of Soldier 
Trail, Santa Catalina Mountains, July 28, 1914, Shreve (GH, MO, NY, US) ; Thornber 
315 (US); April 3, 1894, Tourney (US); R. M. & A. F. Tryon 5052 (B, BM, F, GH, K, 

Mexico, sonora: Meams 2701 (B, US); Wiggins 8330 (US), baja California: 
Oct. 26, 1930, M. E. Jones (BM, MO, US). 




[Vol. 42 


41. Selaginella eremophila Maxon, in Smiths. Misc. Coll. 72 5 :3. 1920. (Holo- 
type: Johnston 104.7 US!. Paratypes: Mearns 3162 US!; Bethel US!; Saunders 
in 1908 US!; Hall US!; Parish 61 II US!; Dudley US). Fig. 59. Map 54. 

Upper leaves fleshy, lanceolate-triangular to ligulate-triangular; under leaves 
papyraceous, to herbaceous at the apex, ligulate-lanceolate to lanceolate-ovate, 
about as long as to half again as long as the upper on the same portion of the stem; 
leaves setate in the bud, setae entirely deciduous to casually persistent, delicate, 
filiform, tortuous. Megaspores rugose-reticulate with irregularly projecting rugae 
on the commissural face, rugose-reticulate on the outer face, pale yellow. 

The tortuous, delicate and early deciduous setae set this species off from all 
others. The setae are attached only on the young leaves in the apical bud and are 
so inconspicuous that they had been overlooked until C. A. Weatherby 24 noted 
them in 1943. 

Jaeger in 1934, Arizona, is unusual in that the setae are only slightly tortuous 
and less delicate than normal. 

Open rocky and sandy deserts, in sheltered places in rock crevices or in sand 
or gravel at the base of boulders, 130-1000 m. 

Southwestern Arizona, southern California, Baja California. 

Representative specimens: 

United States. Arizona: Dec. 17, 1934, Jaeger (US). California: April 17, 1918, 
Bethel (US) ; Dec. 25, 1903, Dudley (US) ; March, 1919, Hall (US) ; Palm Canyon (Palm 
Springs), Riverside Co., Johnston 104.7 (US); Mearns 3162 (NY, US); Parish 1200 
(NY); Rose 45342 (MO); 1903, Saunders (NY), 1908, Saunders (US); R. M. 6 A. F. 
Tryon 5057 (B, BM, F, GH, MO, NY, P, UC, US) ; Wolf 8461 (GH). 

Mexico, baja California: Wiggins 9981 (US). 

42. Selaginella Parishii Underw. in Bull. Torr. Bot. Club 33:202. 1906. (Holo- 
type: Palmer 306 NY. Parish 1 200 and Saunders in 1903, also cited by 
Underw., were excluded by Maxon, in Smiths. Misc. Coll. 72 5 :4. 1920; they 

are S. eremophila. 

Figs. 60, 61. Map 55 

Upper leaves fleshy, lanceolate to lanceolate-triangular; under leaves papy- 
raceous, lanceolate-ovate, up to about twice as long as the upper on the same 
portion of the stem; leaves muticous, the slightly modified apex acute to obtuse. 
Megaspores rugose-reticulate with irregularly projecting rugae on the commissural 
face, rugose-reticulate on the outer face, pale yellow. 

The muticous leaves are sufficient to set this species off from the others of the 
series. In a few specimens some strobili are dorsiventral. 

Crevices of sandstone or slate rocks, 1700-2300 m. 

Coahuila and adjacent Zacatecas. 

Specimens examined: 

Mexico, coahuila: Nil 105 (Arsene 3453) (US); June, 1909, Nil (US); Pennell 
17272 (US), zacatecas: Chaff ey 58 (US); near Concepcion del Oro, £. Palmer 306 in 
1904 (F, GH, MO, US). 

24 Amcr. Fern Jour. 3 3:115. 1943. 



43. Selaginella Landii Greenm. & Pfeiff. in Ann. Mo. Bot. Gard. 5:205. 1918. 
(Holotype: Barms & Land 2024 MO!). Figs. 62, 63. Map 56. 

Upper leaves fleshy, long-deltoid, or with parallel sides toward the base; under 
leaves papyraceous, lanceolate-ovate to ovate, acuminate, more than twice to 2 l /z 
times as long as the upper on the same portion of the stem; leaves muticous to 
short-setate, the upper muticous with the slightly modified apex obtuse, the under 
muticous or usually with a short, stout seta up to 0.3 mm. long. Megaspores 
slightly and finely rugose on the commissural face, slightly rugose, granular, to 
nearly smooth on the outer face, yellow-orange. 

This is the most specialized species of the series in that it is the most strongly 
dorsiventral. The short-setate under leaves and the muticous upper leaves are 
sufficient to characterize it. 

Dry rocks and on boulders, 1700-4000 m. 

Nayarit to Puebla. 

Specimens examined: 

Mexico, nayarit: M. £. Jones 23495 (MO, US), jalisco: Barnes & Land I $3 (F), 
San Esteban Mountains, ca. 32 km. from Guadalajara, 1908, Barnes & Land 2024 (GH, 
MO, US); Vringle 10823 (GH, US); Rose tf Painter 7499 (US), morelos: Matuda 

26lZ2 l\ACS TTQ\ mrcmA. Kr„~sv*.*~ or /TTC\ 

26352 (MO, US), puebla: Kenoyer 25 (US). 


Lycopodium struthioloides Nutt. ex Baker, Handb. Fern Allies, 35. 1887, in synon., 
not Pr. 1825. The brief description is not sufficient to place this name although 
it suggests S, Watsonii. 

Selaginella rupestris (L.) Spring var. borealis Spring, in Nouv. Mem. Acad. Roy. 
Belg. (Monog. Fam. Lycopod.) 24:57. 1850. Indirectly based on a variety 

of specimens representing many species. 

pestris var. brevip 

1865, in 





Selaginella rupestris var. longipila Fourn. Mex. PL 1:146. 1872, nomen nudum. 

Selaginella rupestris var. longipila subvar. glaucina Fourn. Mex. PL 1:146. 1872, 
nomen nudum. Wright 1820 is S. Sellowiu The other specimens cited un- 
doubtedly represent other species. 

Selaginella rupestris var. longipila subvar. viridis Fourn. Mex. PL 1:146. 

nomen nudum. Bourgeau 2541 and Bottero 78 are S. extensa; the other speci- 
mens cited probably represent other species. 

Selaginella rupestris var. tropica Spring, in Nouv. Mem. Acad. Roy. Belg. (Monog. 
Fam. Lycopod.) 24:57. 1850. Indirectly based on a variety of specimens 
representing many species. 

[Vol. 42 



Selaginella struthioloides (Presl) Underw. Bull. Torr. Bot. Club 25:132. 1898. 
(Lycopodium struthioloides Presl, Rel. Haenk. 1:82. 1825.) 

Maxon (Amer. Fern Jour. 11:35-36. 1921) has discussed the misapplication 
of Presl's name by Underwood to Selaginella oregana. In addition to the char- 
acters of size mentioned by Maxon, I may add other reasons supporting his 
position. First, Selaginella oregana is not known to grow on Vancouver Island; 
the nearest station for it is about 160 miles south of Nootka Sound, the locality 
cited for Lycopodium struthioloides. Second, Presl would certainly have placed 
a specimen of Selaginella oregana y with its definite strobilus, under his heading 
"Spicis sessilibus" rather than under "Capsulis axillaribus" where he placed 
Lycopodium struthioloides and other species without strobili. Finally, Greville 
and Hooker, in Bot. Misc. 3:105. 1833, say of Lycopodium struthioloides 
Presl: "We have examined Haenke's specimen, named by Presl himself, in the 
Lambertian Herbarium; and find it to be in no respect different from L. laxum 
of that author, which also exists in the same collection." Lycopodium laxum 
is placed by Herter, Index Lycopod. 55. 1949, as a synonym of Urostachys 
carinatus (Desv.) Hert. (Lycopodium carinatum Desv.), a species of south- 
eastern Asia. This evidently represents another example of a mixture of 
Haenke's labels. 


1. S. rupincola Underw. 

la. S. X neomexicana Maxon 

2. S. Bigelovii Underw. 

3. S. arenicola Underw. 

3 a. S. arenicola ssp. Riddellii (Van 
Eselt. ) Tryon 


3a-b. Intermediate between 


Riddellii and ssp. arenicola 

3 b. S. arenicola ssp. arenicola 
3b-c. Intermediate between 

arenicola and ssp. 




( Underw. ) Tryon 

4. S. Balansae (A. Br.) Hieron. 

5. S. Weatherbiana Tryon 

6. S. viridissima Weath. 

7. S. tortipila A. Br. 

8. S. Sellowii Hieron. 

9. S. Sartorii Hieron. 

Wightii Hieron. 

a. S. "Wightii var. Wightii 

b. S. Wightii var. Phillip 

11. S. cinerascens A. A. Eaton 

12. S. Arsenei Weath. 

13. S. macrathera Weath. 

14. S. shakotanensis (Franch. ex Takeda) 

Miyabe & Kudo 

15. S. Wallacei Hieron. 

16. S. mutica D. C. Eaton ex Underw. 

16a. S. mutica var. limitanea Weath. 

16a-b. Intermediate between var. 

limitanea and var. mutica 

16b. S. mutica var. mutica . 

17. S. Extensa Underw. 

18. S. Wrightii Hieron. 

19. S. Steyermarkii Alston 

20. S. Hansenii Hieron. 

21. S. carinata Tryon 

22. S. indica (Milde) Tryon 

23. S. njamnjamensis Hieron. 

24. S. caffrorum (Milde) Hieron. 

25. S. echinata Baker 

26. S. nivea Alston 

27. S. proxima Tryon 

28. S. Dregei (Presl) Hieron. 




29. S. Vardei Lev, 

30. S. oregana D. C. Eaton 

31. S. Underwoodii Hieron. 

32. S. rupestris (L.) Spring 

33. S. densa Rydb. 

33 a. S. densa var. scopulorum 

(Maxon) Try on 

33a-b. Intermediate between var. 

scopulorum and var. 

Intermediate between var. 
scopulorum and var. 


33 b. S. densa var. densa 

33b-c. Intermediate between var. 


densa and var. Standleyi 
33c. S. densa var. Standleyi (Maxon) 


34. S. sibirica (Milde) Hieron. 

35. S. Watsonii Underw. 

36. S. utahensis Flowers 

37. S. leucobryoides Maxon 

38. S. asprella Maxon 

39. S. peruviana (Milde) Hieron. 

40. S. arizonica Maxon 

33a-b-c. Intermediate between var. 41. S. eremophila Maxon 

scopulorum, var. densa 
and var. Standleyi 

42. S. Parishii Underw. 

43. S. Landii Greenm. & PfeifF. 


Abrams 3126 (2); 3399 (11); 3402 (2); 
47*6> 54^4 (20); 6296 (15); 6711, 
7355> 74S6 (2); 7723, IOO70 (20). 

Abrams & McGregor 760 (35). 

Abrams & Wiggins 207, 341 ( 2 ) . 

Aguilar 1429 (19). 

Alexander & Kellogg 2062 (2). 

Allard 3570, 4570b, 4610a, 5488, 6475 

Amable 1 38 3 (39). 

Amortegui A6l (9). 

Anderson, A. P. 1894, 1975 (7). 

Anderson, J. P. 

Andrieux 2 (9). 

Anect 166 (39). 



543 (24). 
Antunez & Dekindt 3040 
Archer 6764, 7129 (35). 
Archer & Gehrt 1 18 (8). 

Arechavaleta 472 (8). 






10639 (*); 10640 (i) 

10642 (1); 10643 (12) 
17777 (33 a-b, 33b); 1 78 1 7 (5); 17929 

(5); i 793 o (31); 17931 (5); 17945' 

I7903, 17972, 17973, 17974, 17975, 
17976, 17977, 17978 (31); l798o (5); 
17982, 18049 (31); 180 m (39); 18052 



20174 (39); 20985 (31); 20987, 20988, 
20989, 20990, 20992 (39); 21107 
(33a-b); 2IIl8 (16b); 21120 (33a-b). 


Arsene & Benedict 15872, 15873, 15874, 

15875 (31); IS&7& ( 5 ) ; 15877, 15878, 

1603I, 16033, 16152 (33b); 16370 

(33a-b); 16402 (33b); 16643 (16b); 
16644 (33a); 1 6645, 16838 (39); 16884 
(31); 18053 (3D; 18054 (33a-b) ; 18055 
(33a); 18056 (33a, 33a-b-c). 

Aubineau 25 (31). 

Austin 241 (20). 

Bachmann 9 (28). 

Bacigalupi 1015 (16b); 2750 (15). 

Baker, C. F. 2(31); 16, 2746 (15); 5234 

Baker, M. F. 

1,5 (3b). 

Baldwin 5/J<? (32). 


959 (8). 
(522p (8, 39). 

Bang /// (39). 

Barber 65,68 (15). 

Barclay 1568 (15). 

Barkley 1725 (33a, 33c); 16144M 

Barkley & Diettert 4<?02 (15). 



&rf (32); W355 
10561 (18); J0<?7<5 

(9); 10529 
(1); "Ojp 


(9); 20#d (8). 
Bartley & Pontius 167 (32). 
Bartram J2 (40). 
Beattie 5283 (15). 
Beetle 1 566, 3066 (15). 
Benedict / (3 1 ) ; 3 (5) ; 4, 5, 9, 13, 2285 


[Vol. 42 



Bergman 3 1 56 (32). 

Beyrich 1 98 (32). 

Biltmore Herb. 3432a (32); 3432b (3c). 

Blake, S. F. 544, 746, 916 (2); 303 I, 

4762 (32); 10088 (20); 10227 (35); 

10351 (15); II481A (3c). 

Blomquist 5° (3c). 

Blomquist & Correll 4715, 47 16 (7). 

Bogusch 527 (1). 

Boivin & Gillett 8686 (33b). 
Bolander 451/ (20); 6247 (15). 

Bosman 4657 (28). 

Bottero 78 (17). 

Bottomley 2<!?77 (28). 

Boughton N41 (39). 

Bourgeau 2541 (17). 

Brandegee, T. S. 672 (2). 

Brass J^pd, 74475* 1 447 5 (3c). 

Breitung 4024, 5553 (33b); 60/O (32). 

Brewer 7597 ( 2 ) ; 7726, 2103 (35). 

Brinkman £55 (33b). 

Brown, S. 95 (33c); 72o* (33b). 

Buchholz 2 (39); 4 (16a). 

Buchtien 3123, 4200 (39). 

Bues 608 (8). 

Burkart 809,3308,7128 (8); I0II2( 39); 

70/73 (8). 
Burkart & Troncoso 7/774 (8). 
Bush 774, £35, 2<?Jp, ^7^, 5069, 52JO. 

55*7 (32). 

Butler ^74 (15). 

Butters & Abbe 1 72, 1044 (32). 

Butters & Rosendahl 2136, 2955, 3352, 

3595, 4278 (32). 
Butters, et al. 398 (32); 592 (33a). 

Cain 143 (33b). 
Calder 28 W (32). 

Calder & Billard 2995, 4454, 4^ 01 (34). 
Calder & Savile 7647, 7677 (33b); 7732 
(15); 7945 (33a-b); 8072, 8366, 8540 

(15); 9221 (33b) ; 70p/^, 777J5, 77J2J, 

1 1 484, 77<5J4 (33a). 
Camp 74 (15); IO29 (35); 72J7 (32). 
Camp & Schulz 2525 ( 1 ) . 

Campbell & Canfield 531 (16b). 

Cardenas 162 (39). 

Carlson 7347* (1). 

Castetter 7487, 7488, 7492 (31); 7493, 

7494 (33a-b); 7759 (31); 7947 (1«); 

<?575 (33a-b); 8384 (16b). 
Chaffey 5? (42). 
Chaney 18 1 (32). 

Chapman 3432b (3c). 
Chase, A. 1 584, 1 962 (32). 

Chase, V. H. 318 (32); 7676 (18). 

Cheney 3303 (32). 

Chevalier 6668 (23). 

Clark, O. M. 8968, 9628 (31); 10034 

(40); 7<X>3<5 (31). 
Clausen 3674, 3679 (7). 
Clausen & Trapido 3663 (32). 
Clements, F. E. & E. S. 279 (31); 532 

Clokey ^9, 57^, 5765 (2). 
Clokey & Anderson 6508, 6532 (2). 
Colby 3364 (32). 
Cole 5135 (32). 
Conard 774 (30); 246 (15, 30); 7-930 


Constance & Rollins 998 (15). 

Cook & Gilbert 1 58, 240, 244, 248, 305, 

608 (39). 

Cooper, J. G. 462 (2). 

Cooper, T. 662 (24). 

Copeland 774 (18). 

Core 3869 (32). 

Correll 6169 (3c) ; 6615 (32) ; 6711 (7) ; 
10857 (32); 73627 (6); 73<?04 (16b); 
74994, 15000 (31); 75025, 15035 (la); 
7523c? (3a). 

Correll, D. S. & H. B. 770 5 (32); 7897, 
8012 (7). 

Correll & McFarlin 6234 (3c). 

Cory 888 (1); 4722 (3a); 6878 (40); 
&£# (18); 9077 (39); JO294, 11650 
(3a); 77(5/6 (39); 18310 (18); 30143, 
30331 (39); 30971 (18, 39); 3&?35> 
38438, 39454 (18); 40232 (40); 47099 
(3a); 42743 (18); 44*37 (6); 405^ 

Cottam 3674, 3754, 4458 (35); 564/ 

Coville 727 (2); 186 (20); 850 (30). 
Coville & Funston 707 (2); 628 (37); 

2077 (35). 
Coville & Gilman 777 (37). 
Cowles 43 (33b); 428a, 455 (15); 579 

Cox 290 (33a). 
Crandall 7054 (31). 
Cronquist 4955 (3c); 5647 (15). 
Culbertson 4790 (20). 
Curtiss, A. H. 3796 (32); 6774 (3b). 
Cutler 7007 (40); 2580 (32); 3220 (3a). 



5IO (23). 

'5' (31). 

3579 (15); 5861 (35). 

Darrow 2847 (35); 3637 (16b). 
Darrow & Phillips 2777, 278 1 (31) 








Davis, R. J. 941 (15). 

Deam 2 1 58 4 (32); 58692, 63923, 

Degener & Peiler 1 61 1 4 (33b); 

Dekindt 543 (24). 
Demaree 3251B (32). 
Diehl 2517, 2565 (39). 
Dieterlen 1 17 (24). 
Dixon 303 (16b); 335 (35). 
Dodge, C. K. 335 (32). 
Dombey 14 (39). 
Doty 1222 (15); 1868 (30). 
Drake JOJJ (22). 

Drouet & Cohen 4959 (33a-b); 

Drouet, F. & R. 4051 (2). 

Drouet & Richards 33 1 1 (16b); 

(5); 3780 (1). 

Drouet, et al. 3358, 3359, 3633, 
3669, 3687 (1); 4097 (16b). 

Drushel J 1 128 (32). 

Drummond 352 (3a). 

Dudley 7495 (35). 

Duncan 7867 (3c); <?<% (3a-b); 
(7); 7//po (32). 

Duncan & Smith 9339 (7). 

Duncan, et ah 9084 (7). 

Dusen 14 (8), 

Duthie J727 (22). 

Eames, A. J. 0081 (32). 
Eastwood 6391 (2); 72/(?5 (30); 1239I 
(2); /J2p7 (15); 75^5 (33b); I7775> 

17775* (1). 
Eastwood & Howell JI59 (36); 3524 

(33a) 18808 (2). 
Edwards 4/p (18). 
Egeling 26 19 (28). 
Eggleston 2 1 99, 2905 (32); /odtfp (15); 

17487 (39). 

Ehlers 7274 (32). 

Epling 7033, 7284, 10418 (33a). 

Evers 32650, 32668 (32). 

Ewan 2751 (3 5); 4240 (2) ; JI293 (33b) ; 
1/665, II730, 1 1827 (33a); 720^7 
(33a-b); 12659 (33b); 72<5pp (33a-b-c) ; 

'l?4J>'55<5p(33a); 13267 (Hb); 14048 

01*); 14370 (5); /4P50, 75J02 (33b); 

/547/tf (33a-b); 18377, 18395 (33a); 

/(fyfo (33c); 18586 (33a). 
Exell & Mendonca 169 (28). 
Eyerdam 2^7p<? (39); 25041 ( 8 ) . 

Far well 669 , 8 45b , 927 3 (32). 

Fassett 2076, 2760 (32). 

Fassett & Wilson 1 55 1 3 (32). 

Faurie 52, 1577 (14); ^pd, pd^ (34); 

9895, 10543, 1 31 51 (14). 
Fellows 6p2j (32). 
Fendler 702^ (31); 1 02 5 (33b). 
Ferguson 235 (10a). 
Fernald 12 334 (32). 
Fernald & Collins 303 (32). 
Fernald & Jackson 11972 (32). 
Fernald & Long 8416, 1 2336, 23098(52). 
Fernald & Pease 3062, 3063, 3064, 24818 

Fernald, et al. 14068b (32). 
Ferriss, J. H. 26 ( 1 ) . 
Ferris, R. S. 1 623 (20); 2090, 8470 (2); 

8474 (11). 
Ferris & Duncan 2496 (39); 3588 (31, 

39) ; 3588a, 3589 (39) ; J5p2 (40) ; 35OJ 

Ferris & Duthie 941 (33a). 
Fiebrig 3 322 (39); 4007 ( 8 ) . 
Fischer 62 1, 627 (28). 

Fisher 75<5p (28). 

Flett 931 (15, 33a); 752p, 1532 (34); 

7702, 1931, 2019 (15); 3092 (33a). 
Flowers 3249 (36). 

Fogg I 5041 (32). 

Fosberg K69 (35); (5p7 (2). 

Foster, A. S. 839 (30). 

Foster, M. B. et al. 1 305 (3c). 

Fox, et al. 7775 (3c). 

Friend 10451 (3c). 

Friesner 10208, 1 1 235 (32). 

Frye, T. C. & E. M. 2537 (17); 2983, 

3H9 (12). 

Galathea Exped. 3188 (8). 

Gale 383 (2). 

Garrett 7J97, 4<*5p, 5 6 99, 7*04 (35). 

Gates 75225 (32). 

Gattinger 370 6 (32). 

Gentry 1194 0> 39); &*V, £255 (1). 

Gerard 1943 (28). 

Gibbs 4 (28). 

Gibert 1 323 (8). 

Gillett, J. B. 4681, 468 1 A (10b). 

Gillett, J. M. 6022 (33b); 6046 (15). 

Gilman 25 (31). 

Glaziou 3306 (8). 

Gleason p^07, 94 lS > 9 6l 3 ( J 2). 

Glover & Gilliland 480, 77* O ob )- 


[Vol. 42 


Godfrey GH1657 (28); 398 1 (32); 4677 Haynes 2648, 2649, 27 1 1 (41); 2785, 

Godman 6 (10b). 
Godman, et al. 690 (33a). 
Goldberg 19100 (28). 
Gomez & Sousa 85 (28). 
Goodding 2 (41); 4 (2); 5 (31); 56 

(16b); IOI, 102, 103 (40); 173, 178 

(16b); I** (1); 213 (31); ^5 (33a); 

722 (40); 2444 (1); 6/4/ (39); 6142 

Goodman do> (33b); 1 8 58 (35); 2JJ7 

Goodman & Hitchcock 1 514 (35). 
Gorman 675, 4594 (15). 
Gossweiler ^2p<? (24). 
Graham 7592 (16b); £405, 8 $21 (35); 

70004 (16b). 
Grant, G. B. 1191, 2456 (2) ; 6400 (35) ; 

dd/o* (2); 7/44 (35). 
Grant, G. B. & Wheeler 1 26, 6 1 44 (2). 

Grassl 58 30, 7189 (32). 
Gray., Pi. Exsicc. 6lJ (32); 1305 (3c). 
Greenman 4612 (35); 5026* (32). 
Greenman, J. M. & M. T. 4697, 472 1 

(35); 6182 (33a). 
Greenway 2431 (24). 
Griffith 2871 (22). 

Haberer 1555 (32). 
Hagenah Il6o (32). 
Hall, C. C. 155, 173, 177, 178 (35); 184 

Hall, E. 694 (30). 

Hanna 472 (33a); 7270 (33b); 7425 

Hansen 878 (20); £79 (35). 
Harding 218 (32). 
Harper <5 (3b); 70, 16, 17 (3c); 37 

(3a-b); 215, 34-3 (32); /.too (3a-b); 

1852 (3c); 7654, 76 > O0 (3b); 1957,1987 

(3c) ; 3//<5 (3a-b) ; 3403 (3a, 32) ; 3762 

Harris, B. B. 708 ( 1 8 ) . 
Harris, C C. & S. K. 3329 (15); 5555 

Harrison 9160 (35). 
Harrison & Larsen 7634 (35); 7697 (33a) 
Hartman 56,327 (1). 
Harvey ^o*, J 773, 720> (18); 7454 (39); 

7454* (i); '5<V (is); '722 (1). 

Hassler 7379, 72535 1 / 2 (8). 
Hatch & Wilson 283, 325 (19). 
Haupt 705 (19). 
Hayden, A. 9729 (32). 

3016 (11). 

907 (33b); 1684 (32); 

1790, 2328 (33b). 
Heacock 235 (33a). 

Heddle 7745 (32). 

Heller 214 (32); 5024, 5207 (15); 5466 
(20);7l67 (15); 7276 (2); 7583 (20); 
7610 (2); £773 (20); 8604 (2); £995, 
9745, 936*0 (35); 70775 (20); IIO54 
(35); 776*02 (20); 74707 (3c). 

Heller, A. A. & E. G. 4002 (30). 

Heller & Halbach 706 (32) 



Henderson 595* 

Hendrix 765 (2). 

Henry, J. K. 9080, 9084, 9085, 9086 


Henshaw 797 (2). 

Hermann 1 8 51 (32); 4665 (33a-b-c); 

4962, 4995 (35); 5063 (33a-b); 5594 
(33a); 7534, 762 1, 8745, 8774 (32); 
IOO08 (3c); IOO70 (3b); 70706* (7); 

70726* (3c). 

Herrera, F. L. 3009a (8); 308 4 (39). 

Herter 76773, 7o'o > 5<5, 94884, 99600 (8). 

Hildebrandt 2363 (28). 

Hill, R. T. 75 (18). 

Hinckley 7077 (31); 7755 (16a); 7756 

(31); 7757(39); 7756 (31) ; 2747 (40) ; 



Hinton 3453 ( 9 ) ; 697b , , . 
Hitchcock, A. S. 22546 (39). 
Hitchcock, C. L. & Martin 4688 (33a); 

4705 (15); 736> (15, 33a). 
Hitchcock, C. L. & Muhlick 6*725 (33a); 

8499, 0048, 9099 


Hitchcock, C. L. & Stanford 6780 (40) ; 

7248 (12). 
Hodgdon 6ll (32). 
Hoogstraal & Chase 7235 ( 1 
Hoover 7399 (35); 4770 ( 

5085 (15); 5398 (35). 
Horner 607 (15). 
Hosie, et al. 4, 1 46 (32). 
Hotchkiss 161 4 (32). 
House l6S2, 4707, 

7308 I 

_. 4274, 4307 (32) ; 490*2 
496*7 (15); 70279, 76*726*, 

21482, 22671, 27974, 28004 ( 32 ) • 
Howe, M. A. 999 (30). 
Howe, T. D. 66 (35). 
Howell, J. T. 2834, 2886 A (2); 2946" 

(11); 396*9(37); 4557(15); 4692(20); 

5037 (38); 7600 (15); 73527 (33a); 

16761. 26066 (35} : 




Howell, T. 696 (30). 
Hoyle 520 (23). 

Humbert 2823, 3000, 3538 (25); 
7089US (27); Il6l2 (26); 11778 
12439, 12706, 13338, 14.158 (27) 

Humbert & Swingle 

Jones, M. E. & Eggleston 19596 (2) 
Jorgensen 4073 (8). 

(>SS7, Julio IOO (39). 

(25); Jiirgens 29, 308, 309 ( 8 ) . 

(26); 5581, 5705 (27). 

Hunnewell 8365 (15); <?J&5 (30); 

9479 (7) ; p?6d (32) ; 10262 (7) ; 

(25); 55/5 Kaessner 860 (24). 


Kearney 480 (40); p<fy (32). 
Keck 4458 (35). 
Kellogg SlS 02). 

(36); 74445 (32); 7570? (33b); 15711 Kemp jp (33b); 52 (15). 

(33a); 16236, 18335 (32). 

litis 1516 (32); jp2p (15); 3988 (33b); 

4001a (lie); 4399a, 4478a (}Sz) -,4640a 

litis & Clebsch 5715 (7). 

Ingram 2723 (16a). 

Innes & Warnock 479 ( 1 ) ; dotf (18). 

Jackson, J. K. 1045 (23); 7//6 (24). 




Jameson 426 (8). 

Jennings, O. E. & G. K. 6661, 7362, 7389 

(32); 8173, 941 1 > 9490 (15); 14518 

Jermy 542 (3a); 343 (39). 

Johnston, I. M. 7044 (2); 1047 (41); 

I595> 1807 (38); 1808 (2); 1815 (38); 

1 82 4, 1825, 1S26, 1827, 1828 ( 2 ) ; 2420 

(5); 2421 (5, 31); 2422 (5); 2423 

(16b); 2424, 2425 (31); 3004 (11); 

5005 (2) ; 3870 (31); #7/ (16b) ; 3872 

(5); J*7J, J*74 (33b); 3875 (33a); 

i?7<5 (33c); 5577, 3878, 3879 (33b); 

J&?/ (33a); 3881a (33c); Jo^te (33a); 

jtffetf, 3883 (33c); Jcftfy (33a); 3^4^ 

(33c); 3885 (33a); J&W, j&?7 (33c); 

3887a (33b); jm<* (33c); 3888b, 3889 

(33b); j<ypo (33a); 389 1, 3892, 3^93, 

3894, 3895, 3896, 3897, 3898 (33b); 
3S99 (33c); 3900 (31); jptf/ (33b); 
3901a (33c); 3902 (33a); JpOJ (33c); 

3904 (31); JP04* (33a); 3905 (33c); 
jpod, jpo7, jpotf (33b); 7208 (18); 

<fod7 (13); 8607 (It); S683 (6); 8854, 
9106 (18). 

Johnston & Muller 307 (1); 651, 1192 

Jones, G. N. 5539 (33c); 5<5/tf, 577^, 

6287 (15); 7<?o<52 (32). 
Jones, G. N. & F. F. 10644 (33b). 
Jones, M. E. 485 (18); 798 (5); 724<5 

(35); 3023 (2); 5653b] (16b); 5<?22«g, 
5937 (35); 2J4P5 (43); 2<5oj2 (1); 
27018A (1). 

Kenoyer 25 (43); 1053 (18); 75<5<5 (9); 
7724 (1). 

Kenoyer & Crum 2808 (18); jp/o (17). 

Kew Exsicc. 72 39 (33b). 

Kiener 3435 (33b); 4227, <547P(5); £247, 

10288 (33b); 702<?p (31); 77J2J, 23695 
Killip 7772, 22dd, J/007, 36944 (32). 

Kirkwood 74/5 (15). 

Kirkwood & Severy 1718 (15). 

Knight 5296 (32). 

Knobloch 36 ( 1 ) ; 5044, 5053 (39); 5557 

(la); 5573. 59^, 59J<* ( 1 ) ; 59J<?> 59^9 

Knopf JJJ (2). 

Knowlton, F. H. 82, 103 (33a) ; /4/ (31). 
Krause, A. & A. 53 (34); 702 (33c). 
Kreager 294 (15). 

Lacas 512 (18). 

Larsen 5998 (16b); OOotf (35). 

Lawrence, W. E. 7<?57 (30). 

Lawrence, W. H. JO ( 1 5 ) . 

Learn 78 (32). 

Leeds 76 (32); /4P (1); 375» 378, 379, 
380, 381, 386, 388, 389, 392, 397, 398, 
399 (3c) -,837 02); 3208 (3c). 

Leendertz 935, 2568 (28). 

Lehmann KK97, BT776 (8). 

Leiberg 3222 ( 2 ) ; 5535 (31). 

LeSueur 7/4<5 (1); 1540 (18). 

LeTestu 47P4 (23). 

Liebmann 2062 (8). 

Lindheimer 76 (3a); 77 (39). 

Lindig 1523 (9). 

Lindman A213, A2319 (8). 

Little 4196 (40). 

Lloyd, F. E. (incorrect as C. E.) 20 (18) ; 

504 (1, 39); 505 (1). 
Lloyd & Kirkwood 20, 1 42 (18). 
Lodewyks 420 (32). 
Lorentz 442 (8). 
Lossen 246 (8); 426 (39). 

Ludlow, et al. 4116, 5494, r 4233 (29); 

17039, 18606 (22). 
Lundell, A. A. 56 (1, 39);p<5 (18). 


[Vol. 42 


Lundell, C. A. & A. A. 9037 (3 a) 

Lyon 812 

18870 (8). 

Lyonnet 2562 (1). 

Macbride 265 1 (5); 7878 (2). 

MacDaniels 728 (39). 

MacDougal 389 (15). 

MacLeay 114 (24). 

Macoun 530, J424/ (15); 66445, 69265 
(32); 70400', 70637, tfO^*, 86373, 
86374, 86375, 86376, 86377 (15). 

Macoun & Herriot 70371, 70372 (32); 

70373 (33b). 
Mackensen I (18). 

Mackenzie, H. JO (15). 

Mackenzie, K. K. 20 (33a); 2997, 3691 

MacMillan & Sheldon 2596 (32). 
Maguire J0027 (1); /0055 (40); I&/50 

( 1 6b) ; 1 932 1, 19378, 19873, 21090 (35). 
Maguire & Holmgren 27 1 32 (33a); 22607, 

25797, 25966, 26140 (35). 
Maguire, B. & R. R. 10066 (1). 

Maguire & Piranian 1 1 37 1 (16b). 
Maguire & Richards 13166 (35). 
Maguire, et al. 3962, 3963 (35); 4359 

(33a); //745(3 1 ) 5 '4454, -^7', 14691, 
14729 (35). 

Maire, R. 1 533 (4). 

Make & Watson 1229 (33b). 

Mandon #4, /£?/ (39). 

Marcelline 2905 (31). 

Marie 486 (40). 

Marr 7^2 (15). 

Mason, H. L. 3342a (20); 11278 (35). 

Massey 322p (32). 

Matuda 1846 (19); 26/7<5 (1); 26350 

(9); 26352 (43); 26353 (1). 
Matuda & Gold 23572 (18). 
Maxon 1 45, 6293, 6343 (32); 10833, 

10841, 10843 (3c). 
Maxon & Hay 3410 (19). 
McAllister 2 (18); 2<* (39); 5 (33b). 
McCabe 5655, 6o2o*, 7108 (15). 
McDougal 4<?/ (31). 

McFarlin 9374 (3c); JO260 (3b); 10271 

Mcintosh 7344 (32). 

McKeever 38 (2); 39 (11); 40 (2). 

McKelvey 6/p, 1563 (1); 1859 (3a); 

Mearns 225 (1); 237 (16b); 540, 605 
(1) 52044 (33a) ',2687 (I) ',2701 (40); 
3162 (41); 3589 (33a); 3881,4060(2); 
4274 (33a); 449/, 4736 (33b). 

Meislahn 1 38 (3b). 

Merrill & Wilcox I2l8 (33a). 

Metcalf.F. P. 1400 (32). 

Metcalfe, O. B. 276,711,991 (31); J/72 

Metz 2/22 (3a). 
Meyer, F. G. /o^, 256, 1062 (15); /24J 

(33a); 2/69 (30). 
Meyer, F. G. & Rogers 31 1 2 (9). 
Meyer, T. 2227, 3687 (8). 
Milburge 893 (15). 
Millspaugh 4502 (2). 
Mohr 12 (17). 
Moldenke <?&? (3c). 
Moore, A. H. 199, 537, 1561 (32). 
Moore, H. E. 5314 (1). 
Moore, H. E. & Wood 3699 (17); 3926 

(18); 4219 (1)14613 (9). 
Moore, J. A. & Steyermark 3046 (16a); 

3047 (39); 3196 (16a); 3245, 3745 


Moore, J. W. 117 (32); /22 (33b); 7&7 

(32); /o^/ (33b). 
Moore, J. W. & M. F. 10547 (32). 

Moorhead / (18). 
Moritz 370 (9). 
Morley 650 (39). 
Morris 76 (32). 
Moss, C. E. 18988 (28). 
Moss, E. H. 2151 (33b); 2752^ (33a). 
Moxley 345, 346, 347, 421, 876 (2). 
Moyer 99 (32). 
Moyle 483 (32). 
Mueller, C. H. 8256 (39). 
Muenscher 9619, 9673 (15). 
Muenscher, W. C. & M. W. 14361 (32); 
14526 (39); /5402 (15). 

Mulford 121 (39). 

Muller, C. H. 2638 (18). 

Munro I02IA, 1777, 1883 (22). 

Munz 6726 (2); 7536 (35); 7612 (38); 

9079 (20); O^??, 17165 (38). 
Munz & Eggleston 19626 (2). 
Munz & Harwood 3789 (37). 
Munz & Johnston 8761 (35). 
Munz & Keck 4865 (41). 

7999 (39) ; 2705, 2706 (2) ; 4695 (33a) ; Munz, et al. 4226 (37). 

5102, 5109 ( 2 ) . 
McMurphy 453, 531 (15). 
McVaugh 4577, 5137, 5313 (3c); 7675 

Murdoch 4304 (32); 47^7 (33a). 
Murie 1232 (30). 

(18); 7o , 9o > (16a); 8007 (16b); /0446, Napier 2054 (10b). 

//o*29 (17); /4660 (35). 

Nash 1449 (3c). 




Nelson, A. 1345 (33a, 33b); 2448 (33b); 
6368 (33a); 7811 (33a-b, 33b); 8781 
(33b); 10676 (16b); JI217, 11217a 

Nelson, A. & E. 6368 (33a). 

Nelson, A. & R. A. 1 1 58 (40); 1 448 

( 1 ) J 5765 > 5766, 5782a, 5782b (33a). 
Nelson, J. C. 3876 (15). 
Nielsen 3423 (32). 
Nil 105 (42). 
Norton 1 405 (32). 
Novogablenov 547 (34). 
Nuttall, L. W. 553 (2). 


O'Neill 75<?o (3c). 

Oosting 5 J (32); 3640, 34422 ( 7 ) . 

Orcutt 1089, 1097 (18); 70p£ (9); 
1098a (18); 7/02 (39); JI42 (9); J/J/ 
(18); //5/tf (9); 1183 (39); 755*, 
/5tf<5, /p^ ( 1 ) ; 1989, 2218 (18); 2p74 
(1); 3084, 4268 (59); 5125 (9). 

Ortega 5319 (9). 

Otis /0/5, 1074, H39> IH$> 1149, H50, 

Pease 1 1 5, 1 29 50, 1 7967 (32); 22298 
(33a); 22300 (15); 22^/p (33a); 22499 
(33c); 24rp5 (32). 

Pease & Bean 23334, 23581, 26103, 26118 

Pease & Ogden 25070, 251 14, 25140(52). 
Peattie 899, QOO (7). 

Peck 15996a (15). 

Peebles & Harrison 6984 (1). 

Peebles, et al. 480 (40); 4657 (1). 

Pegler £?2 (28). 

Pennell J6890 (18); 77/p# (9); 17272 
(42); 77470" (39); 77753 (9); J7P47 
(17); /o^d (39); 18178, 1 8646, 1 9533 


Perrier 7/6<?, &?<5/, &?6? (25); ^305, 

8305 (26); 831 I, 8312, 8313, 8324, 

8346 (25); 18702 (26). 
Perry & Myers 715 (32). 
Peter 13394 (28). 
Peterson 74 (39). 
Phelps 65,66 (32). 

Phillips, E. A. 775 ( 1 ) . 

7/57, J/52, 1/53, 1156, 1 1 57, 1 1 58 Phillips, W. S. 22<?o (1); 25^, 2do^ 

(15); 1242 (20); /JOp (30). 
Over 11098 (32); 7/35<? (33b); 13678, 
13679, 13M0, 15467, 15883 (32). 

Palmer, E. 38 in 1908 (1, 39); 92 in 
1885 (1); 306 in 1904 (42); 434 in 
1875 (2); 455 in 1885 (18); 4?5 in 
1906 (1); 554 in 1896 (1, 39); 660 in 
1874 (3c). 

Palmer, E. J. 1322, 2834 (32); 954*> 
(3a); 10146, 1 1 389 (18); JJ^// (3a); 
27399, 27461 (3c); 2^^, 2<?o > p/ (32); 
30543, 30661 (39) ; 3087J (16a) ; 3J<5op 
(32); 31951,31952 (\Sz); 32800 (32); 
555&* (3a); 35561, 36810 (32); 37964 
05); jtfj/o' (3c). 

Palmer, L. J. 235 (34). 

Palmer, fm. & King /tfd, 21 4 (32). 

Parish, S. B. 1200 (41); 472<? (2); 5052 
(35); 61 IO (2);6lll (41); 8513 (40); 

(40); 2848 (51); 2861 (I); 2866 (31); 

2930 (40); 2959, 2981 (31). 
Phillips, W. S. & T. K. 3342 (33b). 
Phillips, W. S. & Pultz 2805 ( 1 ) . 
Phillips, W. S. & Reynolds 2QOO (33a, 


Pickett 231 (15). 

Piper 3524 (15) ; 3<?02 (30) ; 4064 (15); 
<5o5p (33a). 

Plank 14 (32). 

Polunin, O. J47<? (22). 

Porsild, A. E. & R. T. 1 59, 322, 690, 1 305 

Porter 1 121 (33a). 
Porter & Rollins 5819 (33a). 
Pringle 224 (40) ; 3900 (17) ; 70<?23 (43). 
Pringle & Conzatti 13OO (18). 
Pultz & Phillips 27/7, 2720 (1); 2749 


9/54(11); W^ ( 2, ii); 70705 (ii)'; P Tx S ' £* A * 3IS6 °' ,8); *** ° 7); 

191S0 (15); /p/p0 (41). 
Parish, S. B. & W. F. 671 (2). 
Parks, H. B. 37921, 7 3920 (18). 
Parks, H. E. 24274 (30). 
Parks, H. E. & S. T. 24IW, 24119 (30) 
Parks, H. S. 2525 (15). 
Parodi 1 4328 (8). 
Parsons itf (15). 

Payson 2503 (31); 2^/4, 2&?3 (33b). 
Payson, E. B. & L. B. 4037 (3 3a, 33a-b) 

4453 (33a); 4917, 5074 (35). 

^5 (9). 
Purpus, J. A. 720 (9). 

Pyron /Op (3c); /// (3b); 100 (3c). 

Pyron & McVaugh 603 (32); 75/7 (3b) 
7730 (32); 2056, 2526 (7); 3003 (32) 
3048 (3b); 3057 (3a-b); 3056 (3b) 
3<>4'5 (3c); 3'0/ (3b-c); 3™7 (3b) 
3774 (3b-c); 377p (3c); 3'5* (3b) 

3755 (3c); 3175 (3b); 3775*(3b, 3b-c) 

Quer 7 (4). 


[Vol. 42 


Raup 630S, 641 1, 6607, 6929, 8 1 84 (32). 

Raup & Abbe 3910 (33c); 4451, 4608 

Rawlinson 284 (32). 

Reed, F. 6180 (2). 

Reed, F. M. 2244 (2); 2466 (35); 4070 

Reeve IOO (32). 

Rehmann 3924 (24) ; 3949 (24, 28) ; 4005 

(2*)-> 4333, 5576 (28). 
Reko 5280 (1). 

Rendle 50/ (28). 

Reverchon 1 5 (18); I^5<5 (39); 1632 

Richards & Drouet 363 (31). 

Riddell J<5, //p7 (3a). 

Riedel 7 (8). 

Robinson, C. B. p/5 (32). 

Robinson, T. R. & Maxon 1 5, 32 (32). 

Rodgers, L. 404 (7). 

Rodin 4403 (28). 

Rogers, F. A. 5683 (28). 

Rogers, H. T. 763 (15). 

Rolland 6131 (32). 

Rollins 51227 (35). 

Rose, J. N. 2879 (1). 

Rose, J. N. & Fitch 17973 (18); -^7/ 

Rose, J. N. & Painter 6536 (39); 6694 

(1); 6835 (S);7499 (43). 
Rose, J. N. & J. S. 1 1 199 (I). 
Rose, J. N. & Mrs. 18845, 19466, 19468 


Rose, J. N. & Russell 19937 ( 8 ). 

Rose, J. N. et al. 8268 (39) ; 9614, 12857 
(I); 23039 (21). 

Rose, L. S. 34495 > 400l6 (2) ; 45039 (20) ; 

45342 (41). 
Rose-Innes 41026 (18). 
Rose-Innes & Warnock 783 (3a). 
Rosendahl 448 (32); #64 (15); 1074 

(33a); 2<5&2 (32). 
Rosendahl & Butters 4564 (32). 
Rosendahl, et al. 1 51 5 (15). 
Rousseau 26929 (32). 
Rudatis £57 (24). 
Ruiz 98 (39). 
Ruth 104, 706, 749 (32). 
Rydberg 1 184 (33b). 
Rydberg & Bessey 3517, 3518, 3519 (33b). 
Rydberg & Carlton dj/o, dj<5<5, 7/<?J, 

7555 (35). 

Rydberg & Garrett 8666, 8997 (33b). 
Rydberg & Vreeland 6588 (33a); (55«?p 
(33b); 6590 (16b, 33b); <J?p/ (31). 

St. John, H. 8668 (30). 

Sandberg, et al. 54, 54a, 54b (15). 

Scamman 878, 1687, 1081, 21 91 (34); 

2493 (33c); 2790 (33a). 
Schaffner JO (39); II (1); pjj (39). 
Schallert 6123 (30). 
Schantz 5/ (5). 
Schelpe 2960 (28). 
Schroeder 72 (18). 
Schulz 753 (8). 
Scouler 335 (30). 
Sellow 670 (8). 
Senn I 678 (32). 

Sharp 44/2<5 (39); 44168, 45375, 45^4, 

46264 (17). 
Sharp, et al. 50255 (17). 
Sharsmith 2076,3127 (35); JP7J (33a). 
Shaw 398, 902, 1060 (33 a). 
Shear 3834 (33a); ^/<5 (16b). 
Sheldon 1493 (32); (fotfj (15). 
Shoop 553 (32). 

Shreve 5120, 5334, 6743, 9041 (1). 
Sim 6095 (23). 
Slater 2, 3 (la). 
Small & Carter 1013 (3c). 
Small & DeWinkeler P7JJ, 9768 (3c). 
Small & Heller 21 4 (7). 
Small & Matthaus 1 1 627 (3c). 
Small & Wilson 1762 (3c). 
Small, et al. 3349 (3c). 
Smiley 5/5 (35). 
Smith, H. 2394 (29). 
Smith, H. H. 3142a (15). 
Smith, L. B. 4679 (8). 
Smith, R. J. 4 (20); 5 (35). 
Sodiro 47/ J (8). 
Soukup 1885, 3571 (39). 
Soxman 191 (3a); 55/ (la); ^p (3a); 

461 (39). 
Spencer 2<5/, 1 468 ( 2 ) ; /^ril* (41); /<5^<5 

(11); 1697, 1907, 2124 (2). 
Sperry 427 (19); 428 (16a); T7JJ (39); 

T738 (i); /M (18). 

Steele & Steele 166 (32). 

Steinbach 8254, 8574 (8). 

Stewart 2204 (6). 

Stewart & Johnston 2122 (1). 

Steyermark 73d, 77P, 1 5644, 15809, 20198, 

27499, 27855, 40256 (32) ; 47 1 1 5, 50501, 

51292 (19); 64653 (32). 
Stanford, et al. 118 (39) ; 2/^7 (9). 
Standley 4153 (33b); 4/5* (5, 31); 455* 

(5); 4<52<5 (5, 31); 5199 (16b); Jjdjtf, 

J^ (31); 14692 (5); 15/73 (33b); 

I53IS (33b); TJj/o'tf (33a); J^J^O, 

J 5345 (15) ; J55<5?( 3 3c); /jjpd, 15752, 




15598, 16216, 16255, 16288, 16378 Tryon, R. M. 1902, 2685, 3439, 4781, 

(33a); 16583 (15); 16970 (33c); 17055 
(33a); 17164(15); 17164a (33a); 17228 
(33c); 17321 (33b); 17483 (33c); 
77774 05); /79/^> 18050 (3 3 a) ; 18 1 36 

4826 (32); 5008, 5009, 5010, 501 I, 
5012, 5013 (3c) ; 5014, 5016, 5017 (3b) ; 
5018, 5019, 5021, 5024, 5026, 5027, 
5093 (3c). 

(33c); 18185, 18606 (33a); 40J<%>(18); Tryon, R. M. & A. F. 5005 (32); 50/5 

52533 (3c); 63702 (19); 73579, 92638 

Standley & Bolman 10830 (31); 77963 

Stork & Horton 9388, 10719 (39). 

Suksdorf 4 (15); 47, 2.25 (33b); 20S 
05) ; 753 (20) ; 1 3 64, 2085, 2644, 2645, 
2646 (15); 6735, 6736, 68 1 1 (30); 
7412,8563 (15); 8834 (33a). 

Summers 2209 (30). 

Svenson & Fassett 302 1 (32). 

Swynnerton Ml 7 a (28). 

Tamayo 2p<? (8); 1753, 1 77 3 (9). 
Tarleton 188 (34). 
Taylor, B. C. 1573 (32). 
Taylor, L. E. 2144 (18). 
Taylor, M. 419 (18). 

Taylor, T. M. C. 1009, TI 35> I2 90 (15); 
2014 (33a-b); 2015, 2023 (33a); 2085 
(15); 9794 (33a); 46024 (15). 

Taylor, T. M. C. & Pillsburt 461 18 (15). 

Taylor, T. M. C, et al. 71, 1 44 (32). 

Tharp 1792 (9)1 43Si3> 43 SH (I 8 ); 

44027 (3a). 

(6); 5030 (3a); 5031, 5032 (39); 5033 
(3a); 5034, 5035, 5037 (18); 3039 
(39) ; 5043 (40) ; J<W (39) ; 5048 (40) ; 
5050, 505I (1); 5052 (40); 5053, 
5054, 5056 (2); 5057 (41); 5059 (38); 
3060 (35); 5063, 5067, 5068 (20); 
5070, 5071, 5072, 5073 (16b); 3074 
(31); 5075 (16b); 3076 (31); 5077 

(5); 307* CO; 3<>79 (5); 5080 (39, 

40); 30^1/4 (40); 5080^/2, 5082 (39); 

30<?3 (16b); 5084 (16*); 5088 (31). 
Tryon, R. M. & P. F. 4695 (32). 
Tuerckheim ££44 (19). 
Turner 1612 (33b). 
Tweedy 36/ (33a). 

Uhlig 845 (10b). 

Ule 232 (8). 

Ulke F20 (33c). 

Umbach 3, 7/7, (?3d (33a); 5166 (32). 

Underwood 7333 (3b, 3b-c); 1355" (3b). 

Underwood & Selby 7 (33b); 2 (5); 733 

(33a-b, 33a-c). 
Underwood & Seymour 2805 (32). 

Van Eseltine 240, 501, 502 (32). 

Thompson, J. W. 2070, 2673 (15); 3773 Van Schaack 2686 l / 2 (33c); 2772, 

(33a); 2977 (16b). 

2934 (33b) 


(30); 4002, 4029, 4058, 4102 (15); 

4"3(30); 4^4(15); 4474> 4484O0); 
4844a, 5263, 5478, 5695 (15); 5779 
(33a); 3972 (15); 5932 (30); 6040 
(15); 6241 (30); 6333 (15); 6673 
(33a); 6797, 7458, 8 1 28, 8747 (15); 
9399 (30); 70724 (15); 703<?3 (33a); 
77423, 12752 (30); 73363 (33a). 

Thornber 373 (1, 40). 

Thorne 3703 (3b); 9394 (3c). 

Thurow 7,8 (3a). 

Thwaites 7474 (10a). 

Tisserant 3587 (23). 

Tourney 7 (16b). 

Tovar 33^ (39). 

Townsend 77 (7). 

Tracy, J. P. 9783, 16452 (15). 

Tracy, S. M. 3370 (3b); 7554 (3c). 

Tracy, S. M. & Earle 275 (16a-b). 

Train 377 (35). 

Trask 27 (2). 

Trelease 420 (\);264I (15); 2642 (34). 

Van Schaack & Freytag 

Vargas 3747, I IO 58 (39). 

Venturi 7744, 1 866 (8). 

Victorin <?427 (32). 

Victorin, et al. 4207, 56848 (32). 

Visher 70, 23 (33b). 

Vorhies 79 (16b). 

Walker, E. H. 329, 2969 (32). 

Walker, E. P. 36 5 O^)- 
Walpole 2747 (30). 
Wallich 2188-3 (10a). 
Ward, L. F. 702* (20). 
Ware, R. A. 379 (35). 
Ware, W. A. 520 (20). 
Warnock T25 (39); 479> 614 O); 924 
(18); 20893 (1); 21694 (6). 

Waterfall 4622 (39); 4811,5392 (I6b); 
6358 (39); 6426a (3a); <*090, <*737 

Watson, P. 30OO (28). 
Watson, S. 4^9 (15); 7370 (35). 



[Vol. 42 

Webster 1 3 (15). 

Weintraub 72a (32). 

Welsh 18089 (24). 

Welwitsch 48, 49 (28); 169 (24). 

Wetherill 563 (16b). 

Wheeler, L. C. 2981, 3036, 3122 (33a); 
3158, 3419 (15). 

Wheeler, W. A. 672, 1058 (32). 

Wherry & Adams 2739 (32). 
White J9&? (1). 

Whitford 47 (33b). 

Wiegand & Manning 50 (3c). 

Wiggins ^2/2 (2, 11); 42 J? (11); 6^5 
(40); 6764 (35); ?2p5, 7407 {1)18174, 
8330, 8694 (40); 9937 (2); 9981 (41); 

7/Jc?p (35). 
Wiggins & Rollins 

450 (1); 4*4, 561 


Wight 74, 2<?5, 3194 (10a). 
Wild /2/<?, 73/5 (28). 

Wilkens 1891 (16a); 202 3 A (39); 2^22 

Williams, L. 1395 (33a); /^2pp (19). 
Williams, L. & R. 2J2<5 (33b). 

Williams, R. S. 534 (33b); 1400 (8); 

2^p (39). 

Williamson IOI7, 2317 (32). 

Wilms 18 1 4 (28). 

Wilson, E. H. 5411 (29). 

Wilson, P. 37 (32). 

Winter 88 (32). 

Wolf, C. B. 3976, 5335 (35) ; 8461 (41). 

Wolf, J. 097 (33a). 

Wolff, S. E. 1324, 31 1 5 (3a). 

Wolley-Dod jpj (2). 

Wood, C. E. 3547 (32). 

Wood, C. E. & Clement 7204, 7503 (3c) ; 

7571 (3b). 
Wood, J. M. 1 1950 (28). 

Woodson 1833 (16b); 1834 (33a). 
Wooton J24 (1). 

Wright, C. 828 ( 1 8 ) ; Jo^O ( 8 ) ; 2106 
21 16; 21 1 5 (16b); 2JI<5 (1). 

Young 21 (39). 

Zeller, S. M. & E. B. 7^ (15). 


Accepted names are in Roman type; synonyms and rejected names are in italic; names 

iUDPeneriP ratpannpc ir* it-* rAm>r»T« 

of subgeneric categories are in capitals. 

Arenicolae 17 is 



Eremophilae 17 76 




Ly co podium 

40, 67 

— 8 2 


carinatum _ 







rupestre 64 


struthioloides 8 1 , 

Rupestres 17 58 

Sartorii 17 3 1 

Selaginella (Section) 16 


Selaginella (Subgenus) ... 






~ 34 
arenaria 26 

arenicola 23 

__ 26 
- 26 

ssp. acanthonota 

ssp. arenicola 

ssp. Riddellii I 24 

arizonica 7$ 

Arsenei 4 1 

Aschenbornii 77 

asprella 7 5 

Balansae 27 

Bigelovii 22 

Bolanderi 49 

Bourgeaui 64 

bryoides 40 

caffrorum ___. 53 


__ 53 

carinata 50 


• • 



var. Karwinskyana 20 

var. Neeana . 20 



columbiana .. 67 

Coryi __. 
deflexa . 






densa „_ 66 

var. densa 68 

var. scopulorum 67 

var. Standleyi 71 

denticulata 1 1 

Dregei 57 

var. Bachmanniana 57 

var. Hildebrandtiana 57 

var. Petersiana 57 

var. pretoriensis . 57 

var. Kehmanniana .... 57 

var. Welwitschiana 57 

echinata 54 

Engelmannii 6 8 

eremophila 8 

extensa 47 

Fendleri 62 

florid ana 26 

funiformis 2 6 

gracillima 1 7 

grisea 57 

Hansenii 49 

Haydenii 76 

helvetica 11 

Hintonii 37 

humifusa 26 

indica 5 2 

Landii 81 

lepidophylla 8 

leucobryoides — 74 

longipila _52, 68 

macrathera 4 1 

Mildei 34 

montanensis 43 

montevideensis 34 

mutica 44 

var. Hmitanea 46 

var. mutica 46 

var. texana __„ 46 

mutica X rupincola - 21 

X neomexicana 21 

nivea 5 5 

njamnjamensis ~~ 52 

oregana 61 

Parishii 80 

pauciciliata 34 

peruviana 77 

var. Dombeyana 77 

Phillipsiana 40 

porrecta 37 

Preissiana 1 7 

proxima — . 56 

pumila 17 

py gmaea 1 7 

Quartiniana . 53 

Riddellii „ 24 

Kosendahlii _ 43 

rupestris — _ 64 

f. amazonica ... 34 

f. amurensis _. 71 

f. Bolanderi .. ..... 49 

f. Bourgeaui 64 

f. brasiliensis 34 

f. caffrorum «. 53 

f. Dregei _ 57 

f. Engelmannii ... ~ 70 

f. Hansenii 49 

f . Haydenii — 70 

f. indica 52 

f. longipila _. 68 

i, manchuriensis — 71 

f. montanensis 43 

f. peruviana ... - 77 

f. Schmidtii - 71 

f. sibirica 71 

f. Wallacei ...-_ 43 



Wrightii - 48 

. Balansae 27 

var. acanthonota 




Bolanderi 49 

borealis 8 1 

_ 34 

var. brasiliensis 
brevipita __ 

var. densa 

Fendleri ~. 


.... 81 

..... 43 


... 70 
.. 62 

var. Hansenii . 


var. incurva - 53 

f. abyssinica 53 

f. angolensis 53 

f. capensis - 53 

var. longipila 81 

subvar. glaucina - 81 

subvar. viridis „ - - 81 

._ 47 

var. mexkana 77 

var. montanensis 43 


var. peruviana 

var. recurva 

f. Dregeana — 
f. Welwitschiana 

.- 77 

. .... 57 




[Vol. 42 



rupincola 20 

var. Sartorii 37 

var. shakotanensis 43 

Sherwoodii 29 





Sartorii . 

var. tropica 82 

Wallace! 43 



var. oregonensis 36 

var. venezuelensis 36 

Schmidtii 7 1 

var. Krauseorum 71 

scopulorum 67 

selaginoides 1 6 





Sbeldonii 77 




spinulosa 1 1 

Standleyi 71 

Steyermarkii 49 

stru thioloides 8 2 

tortipila 29 

uliginosa 1 7 

Underwoodii 62 

var. dolichotricha 62 





viridissima 28 


_ 43 

f. columbiana 43 

Watsonii 7 2 



.. 46 
- 28 
. 39 


var. Phillipsiana 40 

var. vetusta 39 


Wightii 39 



Stachygynandrum rupestre 64 




Tetrastichus 1 7 

Urostachys carinatus 82 

Explanation of Plate 

Series Arenicolae 
Fig. 1. S. rupincola, buds at ground level, Tryon tf Tryon 5051 (MO), X 10. 

Fig. 2. 


Fig. 3. S. Bigelovii, branching rhizome and basal portion of aerial stems, Tryon & 

Tryon 5053 (MO), X 1.75. 

Fig. 4. S. arenicola ssp. arenicola, stems with basal buds and rhizophores, Tryon 5016 

(MO), X 1.75. 

Fig. 5. S. arenicola ssp. acanthonota, basal portion of stems, Tryon 5009 (MO), 

X 1.75. 

Fig. 6. S. Balansae, portion of aerial stem, Balansa (P), X 10. 

Fig. 7. S. Weatberbiana, portion of rhizome, Tryon # Tryon 5077 (MO), X 10. 

Fig. 8. S. Weatherbiana, portion of aerial stem, Standley 4558 (MO), X 10. 

Fig. 9. S. viridissima y portion of aerial stem, Stewart 2204 (GH), X 10. 

Fig. 10. S. tortipila, portion of aerial stem, Rugel (US), X 10. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 1 


Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 2 


[Vol. 42, 19551 


Explanation of Plate 


Series Sartorii 

Fig. 11. S. Sellowii, portion of stem, upper side, Venturi 1866 (US), X 10. 

Fig. 12. S. Sartorii, portion of stem, upper side, Liebmann (GH), X 10. 

Fig. 13. S. Wightii var. Wightii, portion of stem, upper side, Ferguson (US), X 10 

Fig. 14. S. cinerascensy portion of stem, upper side, Kimball (GH), X 10. 

Fig. 15. S. Arsenei, portion of stem, upper side, Arsene 1 064.1 (US), X 10. 

Fig. 16. S. macrathera, portion of stem, upper side, Johnston 8067 (GH), X 10. 

Fig. 17, S. shakotanensis, strobilus, Faurie 9895 (MO), X 10. 

Fig. 18. S. Wallace!, portion of stem, upper side, Wallace (NY ex B), X 10. 

Fig. 19. S. mutica var. mutica, portion of stem, upper side, Meehan (YU), X 10. 

Fig. 20. S. externa, portion of stem, upper side, Pringle 3QOO (GH), X 10. 

Fig- 21. S. Wrightii, portion of stem, under side, Try on & Try on 50 3 4 (MO), X 10. 

Fig. 22. S. Steyermarkii, upper leaves, Hatch tf Wilson 325 (US), X 10. 

Fig. 23. S. Steyermarkii, under leaves, the dark areas are colored red, Hatch & Wihon 

325 (US), X 10. 



[Vol. 42, 1955] 

Explanation of Plate 


Series Sartorh 

Fig. 24. 


Fig. 25. S. carinata, stem with branches, under side, Rose et al. 23039 (US), X 1.25 
Fig. 26. S. carinata, portion of stem, upper side, Rose et al. 23039 (US), X 1.25. 

$ Thomson (I 

Fig. 27. 

Fig. 28. S. n)ar 

Fig. 29. S. caffrorum, portion of stem, under side, T. Cooper 662 (NY), X 10. 

Fig. 30. S. echinata, upper leaves from stem apex, Terrier 8313 (BM), X 14. 

Fig. 31. S. echinata, under leaves from stem apex, Perrier 8313 (BM), X 14. 

Fig. 32. S. nivea, portion of stem, upper side, Humbert & Swingle 5518 (US), X 10. 

Fig. 33. S. proxima, portion of stem, upper side, Humbert ef Swingle K70K (US), 

X 10. 

Fig. 34. S. proxima, sporophylls from upper side of strobilus, Humbert & Swingle 

5705 (US), X 10. 

Fig. 35. S. proxima, sporophylls from under side of strobilus, Humbert ef Swingle 

5705 (US), X 10. 


Ann. Mo. Bot. Gard., Vol. 42 % 1955 

Plate 3 


Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 4 


[Vol, 42, 1955 J 


Explanation of Plate 


Series Rupestres 

Fig. 37. S. Vardei, portion of stem, upper side, £. H. Wilson 54.II (P), X 10. 

Pig. 38. S. oregana, portion of stem, Kautz (YU), X 10. 

Pig- 39. S. Underwoodii> portion of stem, under side, Fendler 1024 (MO), X 10. 

Fig. 40. S. Underwoodii, portion of mat, Arsene 17963 (F), X 1. 

Fig. 41. S. rupestris, portion of stem, upper side, Leeds 76 (NY), X 10. 

Fig. 42. S. rupestris, basal portion of strobilus and leaves, /. K. Small on Aug. 18-27, 

1894 (NY), X 10. 

Fig. 43. S. densa var. scopulorum y portion of strobilus, Standley 1 57 32 (US), X 10. 

Fig. 44. S. densa var. densa, portion of stem, upper side, Boivin & Gitlett 8686 (MO), 

X 10. 

Fig. 45. S. densa var. densa, portion of stem, under side, Boivin & Gttlett 8686 (MO), 

X 10. 

Fig. 46. S. densa var. Standleyi, portion of strobilus, Standley 17228 (US), X 10. 


[Vol. 42, 1955] 


Explanation of Plate 

Series Rupestres 

Fig. 47. S. densa var. densa, portion of mat; the dark area to the right is a region of 

dead vegetative branches, Van Schaack d Freytag 2934 (MO), X 1. 

Fig. 48. S. sibirica, portion of stem, upper side, Calder # Billard 4454 (MO), X 10. 

Fig. 49. S. Watsonii, portion of mat, Tryon & Tryon 5060 (MO), X 1. 

Fig. 50. S. Watsonii, portion of stem, upper side, Watson 1 370 (NY), X 10. 

Fig. 51. S. Watsonii, portion of strobilus, Watson 1 370 (NY), X 10. 

Fig. 52. S. utahensis, portion of stem, upper side, Flowers 3249 (MO), X 10. 

Fig. 53. S. leucobryoides, portion of stem, lateral view, Munz & Harwood 3789 (NY) , 

X 10. 

Fig. 54 S. asprella, portion of stem, upper side, Johnston 18 1 5 (US), X 10. 

Ann. Mo. Bot. Gaud., Vol. 42, 1955 

Plate 5 


Ann. Mo. Bor. Gard., Vol. 42, 1955 

Plate 6 


I Vol. 42, 1955 


Explanation of Plate 


Series Eremophilae 

Fig. 55. S. peruviana, tip of stem in dormant state, Tryon 6 Try on 5031 (MO), 

X 10. 

Fig. 56. S. peruviana, portion of stem, under side, Soukup 188 5 (US), X 10. 

Fig. 57. S. arizonica, portion of stem, upper side, W. S. Phillips 2588 (US), X 10. 

Fig. 58. S. arizonica, portion of stem, under side, Shreve (US), X 10. 

Fig. 59. S. eremophila, leaves from apical bud of stem, Johnston J047 (US), X 14. 

Fig. 60. S. Parishii, portion of stem, upper side, and strobilus, £. Palmer 306 (GH) , 

X 10. 


Fig. 62. S. Landii, portion of stem in dormant state, upper side, Pringle 10823 (GH), 

X 10. 

Fig. 63. S. Landii, portion of stem, under side, Pringle 10823 



Pellaea rufa sp. nov., A. F. Tryon. Rhizoma gracile repens dichotomum 
paleis concoloribus rufis elongato-lanceolato-triangularibus acumine filiformi basi 
cordata. Folia 12—30 cm, longa stipite rufo rachide rufa apicem versus flexuosa. 
Laminae elonga to- triangulares bi-tripinnatae segmentis ellipticis vel ovatis vel 
nonnullis ternatis coriaceis nervis immersis. Sporangia brevissime stipitata. Sporae 
pallide luteolo-fuscae triplantae laeves. Typus: Compton 16402 (US), 

Rhizome slender, cord-like, dichotomously branched, long creeping. Scales of 
the rhizome and stipe-base concolorous, ruddy-tan, elongate lanceolate- triangular, 
sparsely dentate, the tip filiform, the base cordate, scales surrounding the meristem 
pinkish or red. Fronds 12-30 cm. long, approximate, the buds paleaceous. Stipe 
and rachis convex or plane on the upper surface, nearly glabrous, ruddy-tan to red 
becoming darker and gray with age, the upper portion of the rachis usually flexuous. 
Blade 10 cm. long and 2 cm. broad to 25 cm. long and 5 cm. broad, bi-tripinnate, 
elongate-triangular, the pinnae ascending at a broad angle to the rachis, the rachises 
somewhat flexuous. Segments 3 mm. long and 2 mm. broad to 10 mm. long and 
6 mm. broad, elliptical or oval (some ternate), retuse, coriaceous, the veins im- 
mersed and obscure, the margin reflexed or revolute, border lutescent, crenulate, 
the young segments reddish! Sporangia with short stalks less than one- fourth the 
capsule length. Spores 64 per sporangium, tetrahedral, pale yellowish- brown, 
essentially smooth. 

Specimens examined: 

Cape Province: Rock crevices, Ngaap Kop, Laingsburg Dist., Nov. 6, 1944, 
R. H. Compton 164.02 (US) ; Same locality, Dec. 1, 1941, R. H. Compton 12619 
(US); Foothills of Witteberg, April 19, 1925, R. H. Compton 2961 (K) ; Rocky 
hillsides among succulents, The Great Karoo, near Matjiesfontein, Jan. 24, 1948, 
Robert /, Rodin 3342 (UC, K, MO) ; Karoo, Groote Fontein, Mr. Dickson, Lady 
Barkly reed. 5/75 (K) ; Locis rupestribus montosis, siccis in dumetris, Kendo, alt. 
3000-4000, 1838, Drege (BM, K, MO, P). 

Additional localities from the Cape Province have been cited by Compton 1 
from — Whitehill Ridge; Klipbank, Beaufort West Div.; and by Alston and Schlepe 2 
from — Sutherland: Klein Roggeveld, Schietfontein. 

This fern of the South African Karroo has had an elusive record for some hun- 
dred years. It is not surprising for the species is strange among African ferns 



^Trans. Roy. Soc. South Africa 19:270. 1931. 
Jour. South African Botany 18:170. 1952. 

*This paper was prepared in connection with a grant from the Bache Fund of the National 
Academy of Sciences for the study oi desert ferns. 

Issued March 24. 1955. 


[Vol. 42 


that they were long considered as a single species. Another member of the genus, 
P. ovata, with a widespread Cordilleran range is more distantly related. Affinities 
of these species are easily seen in the unique, reddish color of the stipe and rachises 
and of the segments particularly on the immature frond. They are likewise ex- 
pressed in the elongate- triangular form and generally tripinnate division of the 
blade, in the elliptical, retuse form of the segments and elongate, basally cordate 
scales. P. and romedae folia of California and P. intermedia of southern Arizona, 
New Mexico, Texas and adjacent Mexico are distinguished from the Chilian and 
African plants in having the rhizome scales with a sclerotic central stipe. The 
rugose spores and compact, multicipital rhizome of P. myrtillifolia distinguish the 
Chilian plant from P. rufa. 

The first collection, made by Drege was included by Kunze 3 under Allosorus 
andromedaefolius along with collections from California and Chile. Hooker 4 
retained these under Pellaea andromedaefolia. The Chilian species was segregated 
in Kuhn's publication on the Mettenius manuscript. 5 . Later authors regarded the 
African record as erroneous and it was excluded until reinstated by Marloth 6 , 
Compton and more recently by Alston and Schlepe as P. andromedaefolia. 

The range of the species is local largely within a radius of 40 miles in the 
yicinity of Whitehill. Several of the collections, including the type, were made 
by Professor Compton whose regard for this area has resulted in an extensive flora 
of the Whitehill District and a Botanical Reserve for the preservation of Karroo 
plants. The species occurs on dry, rocky hillsides at an elevation of 3,000-4,000 
feet under 5-6 inches of rainfall. It is apparently able to make rapid growth 
during the January rains and to tolerate long periods of desiccation. Adaptations 
to this environment in other genera as Mesembryanthemum, Crassula, Helichrysum 
and Stapelia have resulted in the striking xeromorphic flora of the Karroo. 

Related species of Pellaea in Chile and the Southwestern United States exist 
under environmental conditions of essentially the same extremes. The distribution 
of such closely allied species in three distant regions is not a unique one. Such 
relationships particularly between South Africa and South America have been 
treated at considerable length in studies on the origin of the flora of South Africa 
as well as in systematic treatments of several plants and animals as the earthworms 
AcanthodriluSy the rosaceous genus Acaena y and Menodora of the Oleaceae. The 
range of the latter 7 is nearly identical to these species of Pellaea. Evidence from 
these closely related species of ferns supplies additional data on the floristic relation- 
ships of three continents. 

8 Linnaea 10:503. 1836. 

4 Species Filicum 2:149. 1858. 

5 Linnaea 36:85. 1869. 

6 Das Kapland in Wiss. Ergeb. Deutschen Tiefsee-Exped. auf "Valdivia" 2 3 :276. 1908. 
7 Steyermark, in Ann. Mo. Bot. Gard. 19:100. 1932. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 7 







Emeritus Director 
George T. Moore 


Edgar Anderson 
Assistant Director 
T-Ttinw C. Cutler 

Carroll W. Dodge, 


Robert E. Woodson, Jr.. 

Curator of the Herbariua 

Henry N. Andrevs, 


Rolla M. Tryon, 

Assistant Curator of the 

Hugh C. Cutler, 

Curator fuseum of 
Economic Plant- 

George B. Van Schaac^ 

Honorary Curator of Grasses 

Juuan A. Steyermark, 

Honorary Research Associate 

Frederick. G. Mi £R, 


Alice F. Tryon, 

Research Associate 


Research Associate 

Neix C Horner, 

librarian and Editor 
of Publications 

Gerald Ulrici 

Business Manager 


John S. Lehmann 

Daniel K. Catlin 

Second Vice-President 
Eugene Pettus 

Leicester B. Faust 
Dudley French 
Henry Hitchcock 

Robert Brookings 


Richard J. Lock wood 
A. Vessel Shapleigh 
Ethan A. H. Shepuby 


Arthur C Lichtenberger 

Bishop of the Diocese of 

r si 

Ethan A. H. 


t; versity 




Pre t the Aca 

■ ■.,..- - .-' St. Louis 





Raymond R. Tucker 

Gerald I ««- Secretary 

Volume XLII 

Number 2 


of the 



MAY, 1955 


Studies on Asiatic Relatives of Maize Nalinl 

Lichenological Notes on the Flora of the Antarctic Continent and 

the Subantarctic Islands. I-IV - . - • • • * • « 'j-L.*.' i3i_t40 

Carroll W. Dodge and Emanuel D. Kuaolpn «i-iw 

..- ' „ ,-, T?'--hert E F •".• 5 151-152 

Three New Annonaceae from Panama - - • *od«t ^. 

he Botanical Catalogues of Auguste de St. HiIaire Johr D Dwye ; 15Z _ 170 

Comparison of Juniper Populations on an 0«rk Glade^d^ ^ m _ m 

Old Fields 



Entered as »econa-c 

i&c zh 


of the 

Missouri Botanical Garden 

A Quarterly Journal containing Scientific Contributions from the 
Missouri Botanical Garden and the Henry Shaw School of Botany of 

Washington University in affiliation with the Missouri Botanical 


The Annals of the Missouri Botanical Garden appears four times 
during the calendar year: February, May, September, and November. Four 
numbers constitute a volume. 

Subscription Price $10-00 per volume 

Single Numbers 2.50 each 

Contents of previous issues of the Annals of the Missouri Botanical 
Garden are listed in the Agricultural Index, published by the H. W. Wilson 


of the 

Missouri Botanical Garden 

Vol. 42 MAY, 1955 

No. 2 




The five oriental genera of the Tripsaceae — Coix, Trilobachne, Polytoca, Scler- 
achne and Chionachne — are especially worthy of an investigation more detailed 
than has hitherto been accorded them, not only because of their close relationship 
to maize but in their own right. They form a series of primitive forage plants, 
grain crops, and weeds, concerning which very little, aside from taxonomic descrip- 
tions, has appeared in print. As thorough as the study of maize has been, it is not 
improbable that a better knowledge of its Old World relatives will go some way 
towards helping to solve the big problem it still presents, that of its origin. 

Coix is a cereal so ancient that it is mentioned in Vedic literature, was culti- 



Watt infers, judging by the amount of attention the plant received in early 
botanical writings such as those of Pliny, Rumphius, Loureiro, and Gerarde, that 
formerly it was more extensively cultivated. Wester (1920), however, believes 
that though the hard-shelled forms of Coix were known in Europe as far back as 
the time of Pliny, the soft-shelled kinds do not appear to have been known to 
European writers until the seventeenth century. 

The Arabs are responsible for its introduction into Spain and thereby to the 
West in general, and also for the picturesque name by which it is commonly known 
in Europe, Job's Tears (Watt, 1904). 


understood taxa, largely because of the occurrence of numerous intergrading 
forms. Therefore, those who have worked on the genus (Watt, Vallaeys, and 
Mimeur) do not agree too closely about which taxa should be allotted specific rank. 
In Table I are listed in chronological order as many basonyms in the genus Coix, as 

*An investigation carried out in the Henry Shaw School of Botany of Washington University and 
submitted as a thesis in fulfillment of the requirements for the degree of Doctor of Philosophy. 

Corn Co 




[Vol. 42 
















C. Lacryma Jobi L. Sp. PL 972. 1753. 

C. Lacryma Jobi Thunb. FL Jap. 37. 1787. 

C. arundinacea Lam. Encyc. Method. 3:422. 1789. 

C. agrestis Lour. FL Cochin. 2:551. 1790. 

C. Lacryma-Jobi 1. ^r«/w (Lour.) Backer, Handb. FL Java 2:33. 

C. Lacryma-Jobi forma agrestis (Lour.) [incorrectly attributed to 

Backer by] Vallaeys in Bull. Agr. Congo Beige 39:255. 1948. 

C. pendula Salisb. Prod. 28. 1796. 

C. exaltata Jacq. Eclog. Gram. 60. 1813-1820. 

C. gigantea Koen. ex Roxb. Hort. Beng. 66. 1814. 

C. Lachryma-Jobi var. gigantea (Roxb.) Stapf ex Hook. FL Brit. 
Ind. 7:100. 1896. 

C. aquatica Roxb. in FL Ind. 3:571. 1832. 

C. gigantea Koen. ex Roxb. var. aquatica (Roxb.) Watt, Agr. Ledger 
11:198. 1904. 

C. Lacryma-Jobi 4. aquatica (Roxb.) Backer, Handbk. Fl. Java 2:33. 

C. Lacryma-Jobi forma aquatica (Roxb.) [incorrectly attributed to 
Backer by] Vallaeys, I.e. 1948. 

C. chinensis Tod. in Index Sem. Hort. Bot. Pan. [Palermo], 5. 1861. 
C. Ma-yuen Roman, in Bull. Soc. Nat. Acclimat. II, 7:442-444; 1881: 
C. Lacryma-Jobi var. "Ma-yuen" (Roman.) Stapf ex Hook. FL Brit. 
Ind. 7:100. 1896. 











C. Lacryma-Jobi forma "Ma-yuen" (Roman.) [incorrectly attributed 
to Stapf by] Vallaeys, I.e. 254. 

C. Lachryma var. stenocarpa Oliver in Hook. Ic. PL III, 8: pi. 1764. 

C. stenocarpa (Oliver) Balansa in Jour, de Bot. [Paris] 4:77. 1890. 

C. Lacryma-Jobi forma stenocarpa (Oliver) [incorrectly attributed 

to Stapf by] Vallaeys, I.e. 255. 

C. puellarum Balansa, 1. c. 

C* tubnlosa Hack, ex Warb. in Bot. Jahrb. 13:260. 1891. 

C. lingulata Hack, in Oest. Bot. Zeitschr. 41:5. 1891. 

C. Lacryma-Jobi var. monilifer Watt, I.e. 214. 

C. Lacryma-Jobi forma monilifer (Watt) [incorrectly attributed to 

Watt by] Vallaeys, 1. c. 254. 

C. Lacryma-Jobi var. typica Watt, 1. c. 206. 

ouwehandii Koord. in Bull. Jard. Bot. Buitenz. Ill, 1:191. 1918. 
C. palustris Koord. 1. c. 

C. Lacryma-Jobi 3. palustris (Koord.) Backer, I.e. 

Lacryma-Jobi forma palustris (Koord.) [incorrectly attributed to 
Backer by] Vallaeys, I.e. 255. 

C. poilanei Mimeur in Rev. Internat. Bot. Appl. 31:199. 1951. 
C. lacryma-jobi var. major Mimeur, L c. 202. 
C. lacryma-jobi var. minor Mimeur, 1. c. 204. 










• •.* 







• • * 











• • * » 



























*Here spelled Coyx. 

delimited here, as were found. These bear the unmodified numbers 1 to 21. The 
list includes also a selection from the many combinations which have been made 
in this genus. These follow their respective basonyms and bear the same number 
modified by a letter. A number occurring in a column under one of the names, 



Watt, Vallaeys, or Mimeur, indicates that the corresponding author equates the 
opposite name to that bearing the number given. 

The best-known and most widely spread species is Coix Lacryma Jobi L. 
Originally native to southern and eastern Asia, it has today virtually run wild 
throughout the tropical and warm temperate regions of the world. The other 
species have a comparatively narrow distribution. C. gigantea Koen. ex Roxb. is 
found chiefly in the eastern and central parts of India, in Ceylon, Burma, and in 
Japan. Its variety aqua t tea has been reported in Indochina, western and southern 
India, Burma, and Japan. C. ouwehandii Koord. is endemic to Sumatra. Another 
endemic species is the one set up by Mimeur, C poilanei, found in Laos. 

Vallaeys (1948) suggests the Malay Archipelago as the center of origin of 
CoiXy taking into consideration the multiplicity of varieties present there, some of 
which are not found elsewhere in the wild state. Accordng to Burkill (1935), "it 
seems early to have become a cereal of some importance in the hills of Indochina 

and southern China, and it is probable that the beginnings of the cultivated races 
should be sought for in that part of the world." 

As an article of diet, Coix used to be known as being most important during 
times of famine, helping to tide over seasons when the rice crop failed. Its nutritive 
importance, however, has impressed many investigators, and of recent years it has 
been gaining in popularity as a cereal and has come to be more generally cultivated. 
Wester was responsible for popularizing it in the Philippines where it is known as 
Adlay. Since then similar campaigns have been set in motion in Java, Ceylon, the 
Belgian Congo, Guatemala, Nicaragua, Costa-Rica, Bolivia, Peru, Chile and Brazil. 
Analyzed along with certain other famine foods, its energy value was found 
to be the highest of all those studied (Paton and Dunlop, quoted by Watt, 1904). 
Wester (1920) records the results of analysis of the Bukidnon form of ad lay as 
closely approximating wheat in starch and protein content and exceeding it in fat. 
Comparing it with two other great cereals of the world, he says, "with a greater 
protein and fat content than either rice and corn it is a more complete human 
food than either of these cereals." 

An idea of its productiveness is obtained from Pieris' statement (1936) that a 
crop of adlay from an acre of land will provide sufficient food for a family of 
four people for about one year. Duthie (1888), speaking of its value as a fodder, 
says it is largely eaten by cattle and reported to be very fattening. 

Vallaeys (1948) writes of its curious use as "the object of practices and beliefs 
held in magic" and also of the preparation of a refreshing drink from the grain 
which acts as a diuretic and a depuratory. In hospitals in China it is said to form 
a good and cheap substitute for barley in the making of gruel. It is known to be 
used in the elimination of stones from the bladder (Burkill, 1935). A decoction 
of the roots is given to children as a vermifuge (Ridley, 1906). 

It is best known, however, for its use for decorative purposes, in the making 
of bead-baskets, rosaries, and various articles of personal adornment. Seeds of the 
wild forms with the hard, lustrous shells are the ones thus employed, usually those 

[Vol. 42 



belonging to var. stenocarpa, which has elongated, cylindrical false fruits, and 
those of var. typica, with false fruits of the characteristic tear-drop shape. 

This list of the uses of Coix could be appropriately closed with a quotation 
from Vallaeys. Wherever man has lived it is possible to discover vestiges of 
ancient cultures and of plants whose use is lost. Speaking of the regrettable 
tendency to limit the number of products useful as food, he writes, "Cette limita- 
tion, cette 'standardisation' ne peuvent etre considerees comme un progres. Voila 
pourquoi dan certain pays, on regard la coix comme une 'solution* a des problemes 
d'importance capitale d'ordre dietetique, economique et social". 

Both Polytoca and Chionachne are generic names which owe their origin to 
Robert Brown (Bennett, Horsfield & Brown, 183 8). Many species have been 
described for both genera as a result of which the literature has become cluttered 
with synonyms. In "A contribution to the knowledge of the Indian Maydeae" 
(1931), Henrard made an attempt to clarify their taxonomy. 

The following table is based on Henrard's and Pilger's treatments. It is 
constructed in the same way as Table I. 




















Apluda digit ata L. f. Suppl. 434. 1781. 

Polytoca digitata (L. f.) Henr. in Meded. Rijks. Herb. Leid: 67:10. 1931 

Coix heteroclita Roxb. Fl. Ind. 3:572. 1832. 

Polytoca bracteata R. Br. in Bennett & Brown, PI. Javan. Rar. 20. tab. 5. 

Cyathorhachh Wallichiana Nees ex Steud. Syn. Pi. Gram. 1:403. 1854. 

Polytoca Wallichiana (Nees) Benth. in Jour. Linn. Soc. 19:52. 1881. 

Sclerachne cyathopoda Muell. Fragm. Phytog. Austr. 8:116. 1873. 

Polytoca cyathopoda (Muell.) F. M. Bailey, Queensl. Fl. 6:1849. 1902. 

Polytoca macrophylla Benth., 1. c. 1881. 

Chionachne Massii Balansa in Jour, de Bot. [Paris] 4:78. 1890. 

Polytoca Massii (Balansa) Schenck ex Henr. I.e. 67:9. 1931. 

Polytoca javanica Henr. in Blumea 3:241. 1939. 



Coix arundinacea Koenig ex Willd. Sp. Pi. 4:203. 1805. 

Coix barbata Roxb. Fl. Ind. 3:569. 1832. 

Coix Koenigii Spreng. Syst. 1:239. 1825. 

Chionachne Koenigii (Spreng.) Thwaites, Enum. Pi. Zeyl. 3 57. 1864. 

Coix crypsoides C. Muell. in Bot. Zeit. 19:334. 1861. 

Polytoca semiteres Benth. ex Hook. Fl. Brit. Ind. 7:101. 1896. 

Chionachne semiteres (Benth. ex Hook.) Henr. in Meded. Rijks. 

Leid. 67:16. 1931. 
Chionachne biaurata Hack, in Philipp. Jour. Sci. Suppl. 1:263. 
Chionachne Hubbardiana Henr. in Blumea 3:162. 193 8. 































The species of both Polytoca and Chionachne range in distribution from India 

to Australia. 



cyathopoda is confined to northern Australia and Queensland; P. macro phylla 
grows in the Malay Archipelago, New Guinea, and the Louisiade Archipelago; 
P. javanica has been reported only from Java by Henrard. 

Chionachne Koenig 




The only economic use reported for these grasses is as fodder. Chionachne 
Koenigii, which is widely spread over India, is used as a fodder when young 
(Duthie, 1888). 


All the plants used in this study were grown from seed at the Missouri Botanical 
Garden, both outdoors and in the experimental greenhouse. Through the Plant 
Introduction Station, Beltsville, Maryland, seeds of various varieties of Coix were 
obtained from India, Ceylon, the Philippines, Central America, and the Belgian 
Congo, and large numbers of plants were raised. Since Coix forms a polyploid 
series with haploid chromosome numbers of 5, 10 and 20, it was decided to choose 
one representative of each of these numbers for cytological study, as follows: 

(1) A species with a haploid chromosome number of 5, bearing small, hard, 
chalk- white false fruits, kindly procured for our purpose from Cuttack, India, by 
Dr. N. Krishnaswami, Cytogeneticist, Agricultural College, Coimbatore, India. 

(2) "Blue Adlay", a variety of Coix Lacryma Jobi L., with a haploid chromo- 
some number of 10, obtained from Trinidad through the courtesy of Dr. William 
L. Brown, Geneticist, Pioneer Hi-Bred Corn Company, Iowa. 

(3) A species with a haploid chromosome number of 20, bearing greenish- 
brown false fruits, collected on the Western Ghats (the mountain range that ex- 
tends for 800 miles along the southwest and west coast of India), also sent by Dr. 
N. Krishnaswami. 

Specific or varietal names have not been appended to (1) and (3), because 
their morphological characters do not conform exactly to any of the descriptions 
of Coix species met with in the literature. 

One species of Chionachne — Ch. Koenigii (Sprengel) Thwaites (n =. 10), 
sent by Dr. N. Krishnaswami from the Millet Breeding Station, Coimbatore, India, 
is included in this study as is also one species of Polytoca — P. macrophylla Benth. 

(n = 20) 


Full descriptions of the five taxa are given below, and inclusive herbarium 
specimens are being prepared for distribution. 

No members of the genera Sclerachne and Trilobachm were available for cyto- 
logical or morphological study. 

[Vol. 42 


The following is a key to distinguish the three genera: 

I. Female spikelets enclosed in stony covering formed by modified indurated leaf sheath Coix 

II. Female spikelets enclosed in hardened covering chiefly formed by indurated glume I. 

a. Terminal spikes male, lateral spikes mixed or female Volytoca 

b. All spikes with flowers of both sexes Chionachne 

( 1 ) Coix from Cuttack, India. — 

Root-system very much branched with extremely fine roots. Culm erect with 
numerous basal leaves. Leaf -blades linear, very narrow, up to 5 mm. broad, the 
longest leaves attaining a length of about 60 cm.; prominent white midrib; at 
union with sheath a well-marked suture which is narrow in the middle and fans 
out on either side; upper leaf surface sparsely set with hairs each arising from a 
colorless, crateriform gland. Sheaths ribbed, dotted on the outer face with gland- 
ular hairs similar to those of the blade. Inflorescences axillary, on triangular 
peduncles, as many as 8 arising from one node; false fruit small compared to most 
other varieties of Coix y 8 mm. long and 5 mm. in diameter, flattened on one side, 
often constricted towards the top, turbinate, chalk-white, hard. Male portion of 
the inflorescence up to 3.5 cm. long. Glume I 1 cm. long, lanceolate-ovate, 
slightly falcate, emarginate, the margins reflexed, the glume thereby 2-keeled, en- 
closing glume II, the keels winged, one of them throughout, the other only in its 
upper half, the wings serrate. Glume II about the length of glume I, acute, entire, 

margins infolded. 

This description comes fairly close to that given by Mimeur (1951) for her 
species, Coix poilanei y endemic to Laos. The finely branched root-system, the num- 
erous basal, narrow, linear leaves, and especially the dimensions and color of the 
false fruits, coincide with her description of C. poilaneL She says, however, that 
"les feuilles sont en continuite directe avec les gaines dont elles ne se distinguent 
pas", whereas in our plant the blades are separated from the sheaths by a distinct 
suture. The glands on the sheath are colorless and not brown as stated by Mimeur. 
Also, for the male spikelets she reports "une seule fleur par epillet," while the spike- 
lets of our plant exhibit the 2 -flowered condition. In referring to her species, she 

remarks that it is "tres proche de Coix aquatic a 

1 (C. gigantea var. aquatica Watt) . 
:ussion is n = 5 , which is the same 
as that determined for var. aquatica by Mangelsdorf and Reeves (1939). 

(2) "Blue Adlay", Coix Lacryma Jobi L., from Trinidad. 

This is the wild "Coix Lacryma-Jobi proper" of Watt to which he gave the 
varietal name typica and of which he speaks as "the normal form" referring to 

the false fruits. 

Stem erect. Leaf -blades lanceolate, broad, up to 3.5 cm. in width, in length 
variable, up to 56 cm., auriculate, glabrous, "except for the double row of ascending 
teeth along each of the veinlets of the upper surface — a peculiarity that gives the 
texture of the leaf the appearance of being embroidered and makes it backwardly 
hispid" (Watt, 1904). Writing further about this character, Watt says: "all the 




forms of this species are at once recognized by the texture of the leaf and its 
peculiar scabrosities." Ligule membranous; union of blade with sheath marked by 

a broad suture. Sheaths quite smooth. 


11 mm. long, 


7 mm. in diameter, hard, smooth, polished, bluish- white. Male 

florescence up to 3.5 cm. long, spikelets in two's or three's. Glume I 7 mm. long, 

lanceolate-ovate, emarginate, strongly keeled, winged in the upper half, wings 


Glume 11 about the length of glume I, acute, margins infolded, entire. 

Text-fig. 1. Coix poilanci Mimeur; fig. 2 
All about y$ natural size. 

(3) Coix from the Western Ghats, 
Stem perennial, erect. Leaf 

"Blue Adlay"; fig. 3. Coix gigantea Koen; ex Roxb. 

broad and up to 30 


each arising from a crateriform gland; margins serrate. Sheath also w,th crater,- 

[Vol. 42 


form glands bearing hairs, but here the glands of a dark reddish hue; union of 
blade with sheath marked by a broad suture. Ligule membranous; just above 
ligule, at base of leaf, a tuft of stiff hairs similar to those on the rest of the leaf. 
Inflorescences axillary, 2 or more spikes emerging from the sheath of a short leaf, 
the spike emerging first, being borne on a longer peduncle than that of the others; 
peduncles flattened on one side, glabrous; false fruits elongate, pyriform, 1 cm. 
long, 5-6 mm. in diameter, mouth oblique, pale green with brown markings, 
slightly flattened on one side. Male spikes up to 4 cm. long, drooping, spikelets 
mostly in 3's, the lateral ones sessile and the middle one pedicellate, each 2-flowered, 
Glume I 9 mm. long, ovate-oblong, emarginate, the margins reflexed, the glume 
thereby 2-keeled, enclosing glume II, the keels winged throughout, margins of the 
wings serrulate. Glume II winged, acute, margins entire. 

The description above corresponds in general to that given for C. gigantea by 
Roxburgh (1832) and that by Watt (1904). The important differences are 
discussed below: 

1. Roxburgh describes the sheaths as "smooth", and Watt as "quite glabrous 
and eglandular". In the specimens under discussion, however, the sheaths are 
sprinkled with sharp hairs arising from crateriform glands of a dark reddish color. 

2. The leaves are not "narrow" and "linear" as in Watt's description but "about 

I inch broad" as in Roxburgh's. They are evidently subject to variation, for 
Watt speaks of ". . . . Sir J. E. Smith's sheet of C. gigantea . . . . which has gland- 
dotted broad leaves . . 

3. With regard to the false fruits, Watt describes them as "flattened on one 
side, and with 2 furrows longitudinally on the flat face . . . ." While the ones 
described here are flattened on one side, no furrows are present, nor are any furrows 
mentioned by Roxburgh or by Mimeur (1951). 

The name Coix gigantea is retained for the entity used in this study. The un- 
certainty of its correct application is indicated by a question mark. The chromo- 
some number of "2n = 40" tallies with that reported for C. gigantea by E. K. 
Janaki Ammal (Darlington & Janaki Ammal, 1945). 

(4) Polytoca macro phylla Bentham. 

Stem robust, erect. Leaf -blades up to 55 cm. long, between 2.5 and 3.5 cm. 
broad, glabrous, lanceolate, acuminate, cordate at base; margins dentate; midrib 
prominent; union with sheath marked by a broad suture. Sheath glabrous. Ligule 
narrow, membranous. Female spikelets 1 cm. long, variable in glume I which may 
be entire or toothed at summit or have 2 lateral teeth, sometimes 1 . Male spikelets 
in pairs, each pair made up of 1 pedicellate and 1 sessile spikelet. Glume I up to 

I I mm. long, acute, asymmetrical in the pedicellate spikelet, margins infolded with 
a broad wing along the fold on one side, wing serrate, other side merely folded 
slightly, lanceolate-ovate in the sessile spikelet, margins infolded, winged along the 
fold symmetrically on both sides. Glume II laterally compressed, folded along the 





Text-fig. 4. Polytoca macrophylla Bench- About % natural size. 

middle, keeled along the fold in the pedicellate spikelet; keel winged in upper half, 


The chromosome number 



(5) Chionachne Koenigii (Sprengel) Thwaites. 

Stem perennial, erect, freely branching at the base. Leaf-blades about 40 cm. 
long (said to attain a length of 2 feet, Ranga Achariyar, 1921), up to 1.2 cm. 
broad, linear, with a prominent midrib; union with sheath well-marked by a suture. 
Sheaths with scattered tubercle-based hairs at the base, sheath margins hairy. 
Ligule membranous; nodes densely ringed with hairs. Inflorescences axillary, most 
of them clustered at the ends of very long peduncles which are flattened or slightly 
concave on one side; leaf, in the axil of which several peduncles arise, with a sheath 
about 2.5 cm. long, the blade, although shorter than normal, fairly well-developed; 
each spike in a cluster enclosed in its lower portion by a bladeless awned sheath, 
1-1.5 cm. long; 1, sometimes 2, of the lowermost spikelets female, borne in pairs 
of which one spikelet is sessile and the other pedicellate, the pedicel partly fused 
with the rhachis; rhachis pubescent. Female spikelets oblong, 5.5 mm. long, en- 

[V«l. 42 



Fig. 5. Cbionachne Koenigii (Spreng.) Thwaites. About % natural size. 

closed in glume I which becomes shiny and hard at maturity; at the union of the 
edges of the glume a groove through which the rhachis passes, Male s pikelets 2- 
flowered. Glume J up to 8 mm. long, acute, pubescent, the margins infolded, the 
glume winged along the fold on both sides in the sessile, but only on one side in the 
pedicellate, spikelet; wings serrate. Glume II acute, laterally compressed and 
folded along the middle with a pronounced keel along the upper half of the fold 
in the pedicellate spikelet, only slightly keeled in the sessile. 





A number of cytologists have definitely established the existence of the chromo- 


Maydeae, Euchlaena and Zea. Tripsacum is a well-known exception; all the 
and varieties have 18 pairs of chromosomes or multiples of this number. Among 
the Oriental Maydeae the gametic or reduced number most often encountered is 
also 10; this is evident from the following table: 

1. Coix Lacryma J obi L. 

2. Polytoca macrophylla Benth. 

3. C. Lachryma stenocarpa Oliver 

4. Sclerachne punctata Brown 

5. C. Lacryma- J obi var. "Ma-yuen" (Roman.) Stapf 

6. Polytoca barbara Stapf [Chionachne Koenigii 

(Spreng.) Th wakes] 

7. C. aquatica Roxb. 

8. C. gigantea Koen. ex Roxb. 

9. C. poilanei Mimeur (?) 

Kuwada, 1915 
Avdulov, 1931 
Reeves and 

Mangelsdorf, 1935 
Reeves and 

Mangelsdorf, 1935 
Mangelsdorf and 

Reeves, 1939 
Mangelsdorf and 

Reeves, 1939 
Mangelsdorf and 

Reeves, 1939 
E. K. Janaki Ammal, 

Nirodi, 1955 

The numbers in all the above taxa except 3, 4, and 7 have been confirmed in 
this study. C. aquatica (Roxb.) and C. poilanei Mimeur (?) are exceptional in 
having the gametic number 5, which is the lowest found in the Gramineae and 
reported so far only for the unrelated genera Briza and Anthoxanthum and for five 
species from two of the sub-genera of the Sorghastreae. 


For cytological study entire male spikes were fixed in a solution of glacial acetic 
acid and absolute alcohol 1:3. Two days to a week later they were transferred 
to 70 per cent alcohol and stored under refrigeration. Young spikes that had barely 
begun to emerge from the ensheathing spathe were found to be at the right stage 
for meiotic studies. Temporary smears were stained either with aceto-carmine or 
with aceto-orcein. A few were made permanent by using the dry-ice freezing 
technique described by Conger and Fairchild (1953). 

While pachytene smears of Coix were being examined to investigate the possi- 
bilities for making chromomere analyses, stray cells from somatic anther-tissue 
were noticed showing well-stained and well-spread diploid chromosomes. These 
were assumed to be tapetal cells. Proof of the assumption was obtained when 
young anthers fixed in a solution of 3 parts absolute alcohol to 1 of glacial acetic 
acid were embedded in paraffin, sectioned and stained with crystal violet. 

[Vol. 42 


Smears made for the observation of tapetal chromosomes were treated in a 
slightly different way from smears made for meiotic studies. The contents of an 
anther were squeezed out in a drop of stain and well stirred with an iron needle. 
The debris was allowed to remain. After the placing of the cover-slip the slide 
was heated over a spirit-lamp and pressed vigorously between blotters to release 
the tapetal chromosomes from the nuclear membrane and spread them evenly. 


The tapetal cells of Coix, Polytoca and Chionachne go through the process of 
endomitosis. At the time the pollen mother cells are undergoing pachytene the 
tapetal cells have already passed from the uninucleate to the binucleate condition, 
and smears made at this stage show many of them in endoprophase. Speaking of 
endoprophase stages in tomato tapetum, Brown (1949) remarks that they are 
"valuable for a study of somatic chromosome morphology". The same can be said 
of the three genera of grasses dealt with in this paper. Observations on chromo- 
some morphology were based almost entirely on tapetal cells from anther smears. 
At about the middle of endoprophase, the chromosomes reach a particularly favor- 
able degree of contraction and stain deeply enough to be identified individually. 
Sectioned material shows them surrounded by a nuclear membrane but in smearing 
the membrane may be lost. 

Coix poilanei Mimeur (?). — This species, together with Coix gigantea var. 
aquatica Watt, occupies the lowest position in the polyploid series formed by Coix. 
Its somatic chromosome number determined both from root-tips and tapetal cells 
was found to be 10. At very early endoprophase the chromosomes are very much 
attenuated and are up to five times as long as they are at the end of this phase. 
They are very loosely coiled, and though they often overlap they can easily be 
traced throughout their length (pi. 8, fig. 1). 

A characteristic of tapetal chromosomes in Coix is their striking differentation 
into regions of different staining capacity. Extremely lightly staining areas occur 
which are very constant in their location at a particular stage. These stand out in 
good contrast to the deeply staining portions, thus making the chromosomes easy 
to identify. Usually, and especially when they occur at the ends of the chromo- 
somes, these achromatic regions are visibly split, with the daughter chromatids 
often divergent. A point to be noted at this stage is that quite often along the 
achromatic regions the chromomeres are visible, and two of these may be seen side 
by side in each diverging half of a chromosome indicating that it is made up of 
four chromatids (pi. 8, fig. 3A). "Achromatic" as used here is a relative term. 
These regions appear pale in comparison with the remaining portions of the chromo- 
some. The chromomeres in the achromatic regions often stain clearly enough to 
be distinctly visible. 

The five chromosomes of the haploid set have been designated A, B, C, S and D, 
S standing for the satellite chromosome (pi. 8, figs. 2-6). Chromosome A has a 



very long achromatic region at one end, and Chromosome B has shorter achromatic 
regions at both ends. Chromosome C has two achromatic regions, one in each arm 
midway between the end and the centromere. Chromosome S is recognized by 
the presence of the satellite. In its elongated state it also exhibits an achromatic 
region in either arm but this may disappear on contraction. Chromosome D is very 
like Chromosome S but devoid of the satellite and shorter. 

Figures 2—6 of pL 8 show tapetal chromosomes in varying stages of contrac- 
tion. Even when quite strongly contracted they can be told apart — Chromosome 
A and B by the difference in number and location of the achromatic regions, 
Chromosome S by the satellite, Chromosomes C and D by their difference in length, 
Chromosome D being shorter. When the achromatic regions do not occur at the 
ends of the chromosomes their contraction leaves wide gaps in the body of the 
chromosomes. Figure 6 of pi. 8 shows a late endometaphase with most of the 
chromosomes divided throughout but held together at the centromeres. The 
diverging arms give them a characteristic cruciform appearance. 

Owing to the fact that there may be two or even three places along the chromo- 
somes that show clear spaces and look like centromeres, the true position of the 
spindle-attachment region was ascertained from this stage and from root-tip 
squashes. Root-tip chromosomes examined at anaphase were found to be bent into 
V*s with equal or slightly unequal arms indicating that all the chromosomes had 
median or sub-median centromeres. Plate 10, fig. 2, shows one large tapetal 
nucleus containing the tetraploid number of chromosomes after the completion of 
one endomitosis. Since the chromosomes are in early endoprophase it is probably 
entering a second cycle. It was not ascertained whether a second endomitosis is 
completed nor what degree of ploidy a tapetal cell is capable of attaining. 

When the chromosome number for Coix aquatica Roxb. was reported by 
Mangelsdorf and Reeves (1939) to be 2n == 10 they observed that "no variation 
in the number was found". In C. poilanei Mimeur (?), though the tapetal cells 
most often showed 10 chromosomes, occasionally cells were found where 12 and 
rarely 1 1 were present. Though the nuclear membrane is ruptured on smearing, it 
is possible to recognize the contents as belonging to a single nucleus. Thus the 
possibility that the extra chromosomes might have been displaced from a neighbor- 
ing nucleus due to the pressure exerted while smearing is eliminated. Besides, when 
pollen mother cells were examined, occasionally 6 bivalents were found at diakinesis; 
fig. 1 of pi. 1 1 shows an MI telophase where 6 chromosomes have been distributed 
to either pole. Due to the way in which the anthers were collected it was not pos- 
sible to say with certainty whether this variation in number was between plants or 

between spikelets of one plant. 

Coix Lacryma Johi L.— The tapetal cells of three kinds of this species were 
smeared to see if varietal differences were manifest in the chromosomes. The first 
was "Blue Adlay", i.e. var. typica from Trinidad (described above). The other 
two were two kinds of var. "Ma-yuen". Cultivated forms of C. Lacryma jobi 
have been placed under this varietal name. Cultivation makes itself evident by 

[Vol. 42 


turning the stony capsule spathe soft and brittle and striated. One of the two 
kinds mentioned came from the Philippines and has chalk-white, brittle, and very 
faintly striated capsular spathes. The shape is that of var. typica. The other has 
dark brown, brittle and strongly striated capsular spathes, somewhat elongated and 
with a mouth larger than usual. For convenience these shall be referred to as var. 
"Ma-yuen (1)" and var. "Ma-yuen (2)". 

A general similarity is obvious in the chromosome sets of these three kinds of 
Coix. On the whole, the chromosomes are smaller than in the 10-chromosome 
species. Another difference is that achromatic regions, when they occur, are usually 

found only towards the ends of the chromosomes. Therefore, the pronounced gaps 
that are left in the body of the chromosomes after a certain degree of contraction 
described in the 10-chromosome species are rarely met with here. 

The attempt to discern homologues among the somatic chromosome sets of 
var. typica, var. "Ma-yuen (1)" and var. "Ma-yuen (2)" was confined to those 
chromosomes provided with pronounced achromatic ends as they helped as markers. 
They have been marked by small letters of the alphabet (pi. 8, figs. 7 and 8; pi. 9, 
fig. 1 ) . The nucleolus chromosomes are very similar in all three, with distal ends 
achromatic and split. In var. typica and var. "Ma-yuen (2)" both nucleolus 
chromosomes in a diploid set possess satellites but in var. "Ma-yuen ( 1 ) " a satellite 
was seen on only one of them. Its homologue is seen to be in connection with the 
nucleolus but no satellite was observed. As in the 10-chromosome Coix, the centro- 
meres are all median or sub-median which agrees with Taylor's observations (1925) 
on root-tip chromosomes. This is fairly clear in the tapetal chromosomes (pi. 8, 
figs. 7 and 8; pi. 9, fig. 1) but much more evident in pi. 11, fig. 7, where meiotic 
chromosomes of var. typica are seen in second metaphase. The chromatids prior 
to separation are held together at the centromere which stands out markedly. 

Coix gigantea Koen. ex Roxb. (?). — Here the chromosomes, in spite of being 
twice as numerous as in the varieties of C. Lacryma ]obi y are no smaller. In fact, 
two pairs approach A and B in size, the two largest chromosome pairs in the 10- 
chromosome Coix. Coix poilanei? (pi. 8, fig. 3), the varieties of C Lacryma J obi 
(pi. 8, figs. 7 and 8), and C. gigantea (?) (pi. 8, fig. 2) all show their respective 
tapetal chromosomes in endoprophase at more or less the same degree of contrac- 
tion. By comparing these figures it can be seen that many of the chromosomes of 
C. gigantea (?) are intermediate in size between those of C. poilanei (?) and of 
the 20-chromosome Coixes. 

The centromeres of C. gigantea chromosomes stand out more clearly than those 
of any of the other kinds described so far. Again, all of them are either median 
or sub-median. This is confirmed by fig. 1 of pi. 13, where the chromosomes are 
in prophase of the second meiotic division. The chromatids are held together only 

at the centromeres and the four diverging arms of each chromosome are seen to 
be approximately equal. 

Due to the much larger number of chromosomes present in C. gigantea? it was 
difficult to obtain cells which showed them spread at an appreciable distance from 

The achromatic ends 



one another. The cell shown in fig. 2 of pi. 9 was stained in aceto-orcein and does 
not show the nucleolus. Diakinesis stages clearly indicate the presence of two 
pairs of nucleolus chromosomes. From observations of carmine-stained tapetal 
cells it was seen that two of the nucleolus chromosomes were similar to the ones 
indicated by arrows (pi. 9, fig. 2). Although only one pair of chromosomes has 
been indicated (by X's) in var. "Ma-yuen (2)" and in C. gigantea? as being prob- 
ably homologous, a general similarity between the genomes of the 20- and the 40- 
chromosome Coixes is very much in evidence. The characteristic achromatic 
regions are found in practically all the chromosomes of C. gigantea? and, as in the 
varieties of C. Lacryma Jobi, are mostly confined to the ends. 
are visibly split and often divergent. 

Occasionally, tapetal cells were noticed which had two extra chromosomes. It 
was not ascertained how often this occurred in the tapetal cells but when studying 
the meiosis, counts were made in a hundred pollen mother cells at diakinesis. Of 
these only two were found with an extra bivalent and one with one bivalent 
missing. The variation seems to be within plants. 

Polytoca macrophylla Benth.— The tapetal chromosomes of this species differ 
from those of Coix in that there is no conspicuous differentiation into chromatic 
and achromatic regions. They stain more or less uniformly, as can be seen in pi. 
10, fig. 3, where they are in endoprophase. Two nucleolus chromosomes with 
satellites can be observed in the same figure. Figure 4 of this plate shows them at 
endometaphase, more condensed and already divided as shown by the longitudinal 
split apparent along each chromosome. Conspicuous constrictions indicate the 
position of the centromeres. Plate 10, fig. 5, shows a late endometaphase. The 
divided halves of each chromosome are seen to diverge and are held together only 

at the centromere. 


fig. 6). Such cruciform configurations due to union only at the centromeres are 
sometimes characteristic of colchicine- treated plants undergoing c-mitoses. Levan 
(1938) describes it in Allium. Brown (1949), speaking of endomitotic cycles m 
the tapetum of tomato, says that in some respects they resemble mitoses upset by 
treatment with colchicine. No cruciform stage, however, nor any ev.dence of 
protracted union at the centromere was observed by him. It is not known whether 
the union at the centromere in Polytoca macrophylla and Coix poilanei? ,s much 
protracted but the cruciform configurations form a definite stage in the endo- 
mitotic cycle of the tapetum in these two instances. 

In the contracted condition at endometaphase (pi. 10, fig. 4) , Polytoca chromo- 
somes bear a likeness to somatic chromosomes of corn treated with paradichloro- 
benzene. This produces effects very similar to those of colchicine, one of which is 
to shorten the chromosomes much beyond their normal length. 

Chionachne Koenigii (Sprengel) Thwaites.-In this species of Chionachne the 
„«~,1 ,I,, nmf «n mM . like those of Polytoca macrophylla, stain more or less uni- 

[Vol. 42 


formly. Plate 9, fig. 3, shows them in endoprophase. The presence of satellites on 
the nucleolus chromosomes is doubtful. What seemed like satellites were only- 
observed in one cell in which the chromosomes were in early endoprophase and 
were thin and long. They were not observed either in root-tip chromosomes or in 
pollen mother cells at diakinesis. The centromeres, though not marked by con- 
spicuous constrictions as in Polytoca, are easy to locate. At endometaphase, the 
chromosomes are very much contracted and the beginnings of a longitudinal split 
can be observed in some of them. 


Observations were made on pachytene smears of Coix, Polytoca and Chionachne. 
The pachytene stage in Coix was found to be markedly different from that of 
Polytoca and Chionachne in that conspicuous knobs were observed in Coix that 
were absent in the other two genera. The part played by chromosome knobs in 
the determination of relationships between maize and its relatives is only too well 
known. Considering this feature of chromosome morphology, it could be assumed 
that of the three genera Coix is the most closely related to the New World Maydeae. 
Longley (1941) pointed out the similarity between the chromosomes of Xea and 
those of Coix. Comparing the bivalents of corn and Coix, he showed that the 
variation in size in the two sets of bivalents was very similar. 

A general similarity seems even more apparent on comparing the pachytene 
stages of a 20-chromosome Coix with those of corn. A glance at fig. 7 of pi. 10, 
however, shows that of the six knobs observable only one is internal and the rest 
terminal, a condition at variance with that in corn where the majority of knobs 
are internal. In this respect, therefore, Coix seems to resemble Euchlaena or 
Tripsacum more than corn (knobs in Euchlaena are both internal and terminal 
and in Tripsacum the majority are terminal). Figure 7 shows a pachytene smear 
from a variety of C. Lacryma ]obi obtained from the Sugar Cane Breeding Station, 
Coimbatore, India. From a study of its gross morphology this plant is classified 
as var. typica. Yet the number of its knobs differs widely from that of var. typica 
from Trinidad ("Blue Adlay"). This variation is quite in accordance with the 
situation in corn. Longley (1939), writing about such variation in knob number 
in corn, observes that it is "frequently very pronounced when plants of varieties 
from different geographical regions are compared". In corn, knob-forming centres 
can exist but not be evident unless a knob is formed at the knob-forming point, 
an occurrence which depends on several factors, one of them being the amount of 
knob material available (Longley, 1939). Figure 6 of pi. 10 shows a single 
chromosome of "Blue Adlay", with an elongated, terminal knob. In appearance 
and position it is exactly like one of the terminal knobs of var. typica from 
Coimbatore. It may well be that the position and number of knob-forming centres 
are similar in the two plants but that knobs make their appearance at more centres 
in var. typica from Coimbatore than in var. typica from Trinidad. The knobs of 



var. typica from Coimbatore vary in appearance from large, somewhat elongated 
ones to smaller, more rounded ones (pi. 10, fig. 7). 

The pachytene chromosomes of C. poilanei? were in too tangled a state to 
allow proper observation. The knobs in C. gigantea? (pi. 12, fig. 1) are small, 
rounded and mostly terminal. No knobs were observed in Polytoca macrophylla 
and Chionachne Koenigii. 

If a chromomere analysis of Coix chromosomes were undertaken it would seem 
best to start with varieties of C. Lacryma Jobi. The pachytene stages are easy to 
handle and the chromosomes spread out well enough to be examined. The chromo- 
some number being the same as in corn, similarities with and differences from corn 
chromosomes would be easier to identify. Only late diplotene stages of C. poilanei?, 
"Blue Adlay" and Polytoca macrophylla were observed. These showed a maximum 
of three chiasmata per bivalent in the two Coixes and up to five in Polytoca (pi. 9, 
fig. 4 ; pi. 1 3 , fig. 3 ) . In C. poilanei? the chromosome pairs could be identified at 
this stage and have been marked in the figure with the same letters used to identify 
tapetal chromosomes. By late diakinesis in C. poilanei (?) (pi. 9, fig. 5), all ex- 
cept the satellite bivalent are completely terminalized and lie side by side or are 
connected at the ends by slender filaments. The bivalent of satellite chromosomes 
is still connected at the satellite end and free at the other, forming a V. All five 
bivalents could easily be identified even in this contracted phase and have been 
indicated by letters as in previous figures of C. poilanei (?). 

In "Blue Adlay", five rings and five V's were observed to make up the ten 
bivalents (pi. 11, fig. 2); the achromatic ends mentioned in describing the tapetal 
chromosomes are still noticeable and are especially clear in the distal end of the 

satellite bivalent. 

Diakinesis in C. gigantea (?) deviates somewhat from the perfectly normal for- 
mation of bivalents exhibited by C. poilanei (?) and "Blue Adlay". One association 
of 4 chromosomes was seen in some pollen mother cells and some showed two rings 
of 4 (pi. 12, fig. 5) but the majority had 20 bivalents. These three kinds of be- 
havior were found in pollen mother cells from the same spikelet. Evidently the 
two sets of 4 chromosomes that tend to form rings possess some incipient homology. 

C. gigantea (?) presumably has two pairs of nucleolus chromosomes. Two 
bivalents were often seen in association with the nucleolus at diakinesis and one 
pollen mother cell was observed which had two distinct nucleoli each associated 
with a bivalent. But the prominent satellites characteristic of Coix were only seen 
to be displayed by one of the bivalents (pi. 12, fig. 6). 

Both Polytoca macrophylla and Chionachne Koenigii behave normally at 
diakinesis, regularly forming 20 and 10 bivalents respectively (pi. 13, fag. 4; pi. 14, 
fig. 3 ) . The former has satellites to its two nucleolus chromosomes but none were 
observed in Chionachne. 

Formation of metaphase plates and separation at anaphase I take place in the 
normal fashion. There is an orderly migration to the poles; and no formation of 

[Vol. 42 

bridges, lagging chromosomes, nor any other irregularities were observed in any of 
the five plants (pi. 9, fig. 6; pi. 11, fig. 3; pi. 12, fig. 7; and pi. 14, fig. 4). 

At telophase I the chromosomes are seen to have formed two compact groups. 
By the time they emerge from telophase I and enter prophase II, the arms of the 
component chromatids of each chromosome are seen to diverge from the centro- 
mere with the result that the chromosomes appear cruciform. A nucleolus has 
been re-formed in each group. This behavior is similar in all five taxa under dis- 
cussion (pi. 12, fig. 3; pi. 11, fig. 6; pi. 13, fig. 1; pi. 14, fig. 1). 

At this stage, the second meiotic prophase, a curious body was seen to occur 
regularly in each of the two groups of chromosomes in C. gigantea (?). It ap- 
peared to be composed of a few darkly staining granules embedded in a lightly 
staining matrix (pi. 14, fig. 1). It varied in shape, being sometimes small and 
rounded, sometimes larger and lens-shaped. The number of granules also varied. 
An attempt to investigate the nature of this body was made by halving an anther 
containing pollen mother cells in second prophase and staining one half with aceto- 
carmine and the other with aceto-orcein. With aceto-carmine both nucleolus and 
the body showed up clearly, while the orcein-stained preparation stained neither 
the nucleolus nor the matrix. The granules showed extremely faintly. It would 
seem from this evidence that the body was in some way connected with the nuc- 
leolus. It was not seen in contact with any of the chromosomes or with the 
nucleolus, and so far as was observed was not seen to be in later stages. 

The remaining stages of the second meiotic division leading to tetrad formation 
are completed in normal fashion. In C. gigantea? and Ch. Koenigii at anaphase II 
separation of chromatids and migration along the spindle may not take place 
simultaneously in both cells (pi. 13, fig. 2; pi. 14, fig. 6). 

Relationship between Chromosome Number and Gross Morphology 

Coix Poilanei?, with 10 somatic chromosomes, is essentially a small plant and 
even when grown outdoors did not exceed a height of three feet. The varieties of 
C. Lacryma-Jobi, 2n = 20, of which several forms of var. "Ma-yuen" and var. 
typica from different regions were raised, showed a wide variation in size, some 
attaining a height of 3 feet, others growing 9 feet tall. Most of them averaged 
7-8 feet. 

Coix gigantea?, with the highest chromosome number (2n = 40), exhibits 
characteristics typical of polyploids. Roxburgh (1832) describes it as attaining 
a height of 8-15 feet. Our plants of C. gigantea? were only between 5 and 6 
feet high, but the leaves were coarser than in the two other species and the plants 
came into flower much later. 


tionship between Chromosome Number and Geographical Distribution 
Species of Coix seem to have reached an optimum in the somatic number of 20 


195 J] 



On finding that accurate chromosome counts could be made from tapetal cells, 
a study of their chromosome morphology was undertaken. While tapetal chromo- 
somes at the right stage for study were being sought it gradually became apparent 
that they undergo what is essentially a process of endomitosis. Endomitosis in 
tapetal cells was first reported by Witkus (1945) in Spinacia. Brown (1949) 
showed that tapetum endomitosis in tomato differed in some respects from the 
process in Spinacia. In both cases the process ends in the formation of polyploid 
nuclei but whereas in Spinacia the nuclear membrane remains intact throughout 
and the chromosomes are irregularly distributed, in tomato there is a breakdown 
of the nuclear membrane and the chromosomes form clumps or are arranged in 

plates at endometaphase. 

In spite of the breakdown of the nuclear membrane, Brown retains the term 
"endomitosis," explaining that "lack of movement of the chromosomes on a spindle 
rather than persistence of a nuclear membrane would seem to be the better criterion 

of endomitosis." 

In the present study the behavior of tapetal cells follows a somewhat similar 
pattern to that described in tomato. The cells are uninucleate to start with but 
soon become binucleate following a normal mitosis, i. e., with spindle formation 
but without the formation of a cell wall. The chromosomes become visible as sep- 
arate entities at endoprophase and contraction commences. At endometaphase the 
chromosomes are very much contracted and the beginning of a longitudinal split, 
prior to anaphase separation, is distinctly noticeable (pi. 8, fig. 9; pi. 10, fig. 4). 
However, the cruciform configurations at late metaphase described in C. poilanei? 
and Polytoca and presumably present in the other three taxa are reported by Brown 

to be absent in tomato. 

At endoanaphase the chromosomes form dense clumps. The fact that the 
chromosomes fall apart at this stage was only gathered from the small size of stray 
chromosomes at the periphery of the clumps. 

That the completion of endomitosis results in tetraploid nuclei was concluded 
from the larger size of many resting nuclei and from instances like the one por- 
trayed in pL 10, fig. 2, where a tetraploid nucleus of C. poilanei} is shown with 
its chromosomes in endoprophase of a second endomitosis. Whether this is con- 
cluded with the formation of an octoploid nucleus is not known Except in C. 
Poilanei} no tetraploid nuclei were observed in a state in which the chromosomes 
could be counted. For this to be possible the chromosomes would have to emerge 
from the resting stage and start on a second endoprophase, and this does not seem 
to happen in the other four taxa, and was only rarely observed in C potlana} 
Presumably then, endomitotic activity stops after one division and the formation 
of tetraploid nuclei, i. e., two tetraploid nuclei per tapetal cell. Brown reports 
three endomitotic cycles as normal to tapetal cells in tomato. . 

Sectioned material at endoprophase showed the nuclear membrane intact and 
fig 9 of pi. 8 shows an endometaphase in van "Ma-yuen ( 1 ) " with the membrane 


[Vol. 42 

still intact. Whether it disappears at a later stage is not definitely known but it 
seems unlikely. Assuming that it does not, the endomitosis in Coix, Polytoca and 
Chionachne resembles Spinacia regarding this point. According to Brown, in 
tomato the membrane disappears toward late endoprophase. 

From what study has been made on Coix chromosomes it would appear that 
the various species and varieties dealt with in this paper form a fairly homogeneous 
group, with 5 as the basic chromosome number. C. poilanei? would then be con- 
sidered a diploid species, C. Lacryma Jobi and its varieties tetraploids, and C. 
gigantea? an octoploid. Morphologically, the octoploid is similar to the diploid in 
an outstanding characteristic, the presence of crateriform glands bearing hairs. 
(Watt considers this characteristic as one of the more important ones that sep- 
arate what he terms "the gigantea- aquatic a series" from the Lacryma-Jobi series). 
In its chromosomes, however, the octoploid is more like the tetraploids. 

Since no crossing experiments were undertaken for this study any hypothesis 
regarding the possible origin of the tetraploid and octoploid should be made with 
the necessary caution. It could be surmised that the tetraploids arose by a crossing 
of two similar diploids and a doubling of the chromosome number of the hybrid. 
They function like true diploids forming no multivalents. In the octoploid two 
nucleoli have been observed, and there are two sets of nucleolus chromosomes. 
Multivalent configurations are found. It probably originated from the tetraploid 
as an autopolyploid. Since its origin, its genomes have undergone changes so that 
it has come to function essentially as a diploid but the occasional formation of 
tetravalents indicates its autopolyploid beginnings. 

On the basis of its pachytene chromosomes, Coix as a genus seems closer to the 
New World Maydeae than to Polytoca and Chionachne. In staining reactions, as 
previously described, Polytoca and Chionachne behave differently from Coix. 
Morphologically, they are similar in having the walls of their fruit-cases 
formed from the hardened first glume, whereas Coix has hardened, modified leaf- 
sheaths. The probability must not be overlooked that investigation of species of 
Polytoca and Chionachne other than those treated here might disclose the presence 
of pachytene knobs. The resemblance of endometaphase chromosomes of Polytoca 
to those of artificially shortened somatic chromosomes of corn has been mentioned. 
On the basis of the present evidence, it would seem that Polytoca and Chionachne 
bear a closer affinity to each other than to either Coix or to the New World 



Avdulov, N. P. (1931). Karyo systematische Untersuchung der Familie Gramineen. Bull. Appl. Bot., 

Genet. & Pi. Breed. (Leningrad), Suppi. 43:1-428. 
Backer, C. A. (1927). Handboek voor de Flora van Java 2:32-33. 
Bailey, F. M. (1902). Queensland Flora 6:1848. 

Balansa, B. (1890). Catalogue des Graminees de PIndo-Chine Franraicp. 

Bennett, J. 

ne Francaise. Jour, de Bot. 4:76-84. 
Plantae Javanicae Rariores. pp. 15-19. 



Bentham, George (1878). Flora Australiensis 7:515-516. 

, (1881). Notes on Gramineae. Jour. Linn. Soc. Bot. 19:14-134. 

•, and J. D. Hooker (1883). Genera Plantarum 3:1112-1115. 

Berger, C. A., and E. R. Witkus (1943). A cytological study of c-mitosis in the plant Spinacia 

oleracea with comparative observations on Allium cepa. Bull. Torr. Bot. Club 70:457-467. 
Bor, N. L. (1940). Flora of Assam 5:453-459. 

Brown, Spencer W. (1949). Endomitosis in the tapetum of tomato. Amer. Jour. Bot. 36:703-716. 
Burkill, I. H. (1935). A dictionary of economic products of the Malay Peninsula 1:629-631. 
Conger, Alan D., and Lucile M. Fairchild (1953). A quick-freeze method for making smear slides 

permanent. Stain Technol. 28:281-283, 
Cooper, D. C. (1933). Nuclear divisions in the tapetal cells of certain angiosperms. Amer. Jour. 

Bot. 20:35 8-364. 
Darlington, C. D., and E. K. Janaki Ammal (1945). Chromosome atlas of cultivated plants. 

397 pp. London. 
Duthie, J. F. (1888). The fodder grasses of northern India. 90 pp. Roorkee. 
Gamble, J. S. (1934). Flora of Madras 3:1705. 
Hackel, E. (1891). Descriptiones Graminum novorum. Oest. Bot. Zeitschr. 41:5—9. 

, (1906). Notes on Philippine II. Philipp. Jour. Sci. Suppl. 1:263—269. 

Henrard, J. Th. (1931). A contribution to the knowledge of the Indian Maydeae. Meded. Rijks 

Herb. Leiden 67:1-17. 

(1938). On a new Chionachne from Queensland. Blumea 3:23 8-242. 
(1938). On a new species of Polytoca from Java. Ibid. 238-242. 

Heyne, K. (1927). Die nuttige planten van Nederlandsche Indie 1:150-152. 

Holland, T. H. (1926). Adlay — its uses and prospects. Dept. Agr. Ceylon, Yearbook 1926:60-61. 

Hooker, J. D. (1896). Flora of British India 7:100. 

Jacquin, J. F. von (1820). Eclogae Graminum. p. 60. 

Koorders, S. H. (1918). Beschreibung einer von Dr. Ouwehand im Toba-See, in Sumatra, entdeckten 

neuen Art von Coix. Jard. Bot. Buitenzorg, Bull. Ill, 1:190-191. 
Kuwada, Y. (1915). Ueber die Chromosomenzahl von Zea Mays L. Bot. Mag. Tokyo 29:83-89. 
Lamarck, J. B., A. P. M. de (1789). Encyclopedic Methodique. 3:422. 

Levan, Albert (1938). The effect of colchicine on root mitoses in Allium. Hereditas 24:471-486. 
Linnaeus, C. (1753). Species plantarum. pp. 972-973. 

Linnaeus, filius (1781). Suppl. PI. Syst., Veg., Gen. PL, et Sp. Pi. p. 434. 
Longley, A. E. (1939). Knob positions on corn chromosomes. Jour. Agr. Res. 59:475-490. 

i (1941). Chromosome morphology in maize and its relatives. Bot. Rev. 7:263-289. 

Loureiro, J. de (1790). Flora Cochinchinensis 2:550-551. 

Mangelsdorf, P. C, and R. G. Reeves (1939). The origin of Indian corn and its relatives. Texas 

Agr. Exp. Sta. Bull. 574:1-315. 
Merrill, E. D. (1906). An enumeration of Philippine Gramineae with keys to genera and species. 

Philipp. Jour. Sci. Suppl. 1:320-321. 
Mimeur, Genevieve (1951). Systematique specifique du genre Coix et systematique varietale de 
Coix Lacryma-J obi. Morphologie de cette petite cereale et etude de sa plantule. Rev. Internat. 

Bot. Appl. Agric. Trop. 31:197-211. 
Miiller, Carolus (1861). De graminibus novis vel minus cognitis. Bot. Zeit. 19:332-335. 
Mueller, F. von (1875). Fragmenta Phytographiae Australiae 8:116-117. 
Pieris, H. A. (1936). Adlay. Trop. Agr. (Ceylon) 86:217-219. 
Pilger, R. (1940). Gramineae. III. Unterfamilie Panicoideae. Engler & Plantl's Nat. Pflanzenfam. 

14 e :184-201. 

Ranga Achariyar (1921). A handbook of some South Indian grasses, pp. 140-142. 

Reeves, R. G., and P. C Mangelsdorf (1935). Chromosome numbers in relatives of Zea Mays L. 

Amer. Nat. 69:633-635. 
Ridley, H. N. (1906). Malay drugs. Agr. Bull. Straits & Fed. Malay States 5:193-206. 

Roxburgh, W. (1832). Flora Indica 3:568-572. 

Royen, A. Van (1740). Flora Leydensis. p. 72. 

Salisbury, R. A. (1796). Prodromus stirpium in Horto ad Chapel Allerton vigentium. p. 28. 

Smith, F. H. (1933). Nuclear divisions in the tapetal cells of Galtonia candicans. Amer. Jour. Bot. 

Sprengel, Curt. (1825). Systema vegetabilium 1:238-239. 
Stapf, O. (1SSS). Job's Tears (Coix Lachryma L. var. stenocarpa). Kew Bull. Misc. Inf; 

Steudel, E. G. (1854). Synopsis Plantae Graminum 1:403. 
Taylor, William Randolph (1925). Chromosome constrictions as distinguishing characters in plants. 

Amer. Jour. Bot. 12:238-244. 

[Vol. 42, 1955] 


Thunberg, C P. (1784). Flora Japonica. p. 37. 

Thwaites, G. H. K., and J. D. Hooker (1864). Enumeratio plantarum zeylaniac. p. 357. 

Vallacys, G. (1948). Le "Coix Lacryma-Jobi*\ Bull. Agric. Congo Beige 39:247-304. 

Warburg, O. von (1891). Beitrage zur Kenntnis der papuanischen Flora. Bot. Jahrb. 13:230-272. 

Watt, George (1889). A dictionary of the economic products of India. 2:492-500. 

, {1 904). Coix spp. or Job's Tears. A review of all available information. Agr. Ledg. 

Weatherwax, P. (1926). Comparative morphology of the Oriental Maydeae. Indiana Univ. Studies 

Wester, P. J. (1920). Notes on Adlay. Philipp. Agric. Rev. 13:217-222. 

Willdenow, C. L. (1805). Species Plantarum 4:202-203. 

Witkus, E. R. (1945). Endomitotic tapetal cell divisions in Spinacia. Amer. Jour. Bot. 32:326-330. 

Explanation of Plate 


Figs. 1—4. Coix poilanei Mimeur. Tapetal chromosomes in endoprophase. 

Fig. 5. Coix poilanei. Tapetal chromosomes in endometaphase. 

Fig. 6. Coix poilanei. Tapetal chromosomes in late endometaphase. 

Fig. 7. "Blue Adlay" (C. Lacryma Jobi var. typica Watt), from Trinidad. Tapetal 
chromosomes in endoprophase. 

Fig. 8. Var. "Ma-yuen (1)" (C. Lacryma Jobi L. var. "Ma-yuen" Stapf). Tapetal 
chromosomes in endoprophase. 

Fig 9. Var. "Mayuen (1)". Tapetal chromosomes in endometaphase. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 







.;■.■ • ^ 


. .. 







« A 





^*«\. v 










- * 











B £ 





• - 




a *r% 

I - 


[Vol. 42, 1955] 


Explanation of Plate 


Fig. 1. Var. "Ma-yuen (2)" (C. Lacryma-Jobi L. var. "Ma-yuen" Stapf). Tapetal 
chromosomes in endoprophase. 

Fig. 2. Coix gigantea Koen. ex Roxb. Tapetal chromosomes in endoprophase. 

Fig. 3. Chionachne Koenigii (Spreng.) Thwaites. Tapetal chromosomes in endo- 

Fig. 4. Coix poilanei. Mimeur. Diplotene. 

Fig. 5. Coix poilanei. Diakinesis. 

■ * 

Fig. 6. Coix poilanei. Anaphase I. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 9 

^ * 

> .>■-- 

a ^ / t 


~ i ^5 '■ 


'••*■ >«& 

> - fc • • ■ - ■ ■■-" 



>- f ■ 2ft 

* -■ ■ 

*s & ■ .* 


s *g 



■a ■ 

4%v ,,,, 

• . 

r, •- -; 





' ■ 

. . ■• 5; <**' '-r .<-.- T 

- . * - 


* -^ ... * ^ 




^ - * *t 



• vv 






*€ pi 


Jtts. £' 


■ - 



; ^' 





. ** 


, 3 'c' ■■'■■;'■■ 




'^T ;.. v . ^.. 

D % 

, v-; 

. MM *-.■. ■ 


* t 





* . . ■ 





[Vol. 42, 1955 J 


Explanation of Plate 


Fig. 1. Coix poilanei Mimeur. Showing twelve tapetal chromosomes in endoprophase. 

Fig. 2. Coix poilanei. Tapetal cell with tetraploid number of chromosomes under- 
going a second endomitotic cycle. 

Fig. 3. Poly toca macrophylla Benth. Tapetal chromosomes in endoprophase. X 13 50. 

Fig. 4. Poly toca macrophylla. Tapetal chromosomes in endometaphase. X 13 50. 

Fig. 5. Poly toca macrophylla. Tapetal chromosomes in late endometaphase. 

Fig. 6. "Blue Adlay." Pachytene. X 13 50. 

Fig. 7. Coix Lacryma J obi var. typica, from Coimbatore. Pachytene. X 13 50. 

Ann. Mo. Bot. Gard., Vol. 42, 195 5 

Plate 10 

— ,■.. Miiiiu^w^iinigp.«spw 






S 1 











** MB 





^MMi * 



Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 1 1 




• ■-■■-*,%.•* 

t« ■ ■ 



^ « 



1 I 







rife £#> / 




& - 


7 ,- 

















»* r 


Y ff? 











I Vol. 42, 1955) 



Explanation of Plate 


Fig. 1. Coix poilanei 

Fig. 2. "Blue Adlay 

Fig. 3. "Blue Adlay 

Fig. 4. "Blue Adlay 

Fig. 5. "Blue Adlay 

Fig. 6. "Blue Adlay 

Fig. 7. "Blue Adlay 

Mimeur. Telophase I of a 12 -chromosome cell 



" Anaphase I. 
M Telophase I. 


Late Telophase I 

" Prophase II. 
M Metaphase II 

IVol 42, 19551 


Explanation of Plate 


Fig. 1. Coix gigantea Koen. ex Roxb. Pachytene. X 13 50. 

Fig. 2. Coix poilanei Mimeur. Telophase I of normal 10-chromosome cell. 

Fig. 3. Coix poilanei. Prophase II. X 1150. 

Fig. 4. "Blue Adlay." Telophase II. 

Fig. 5. Coix gigantea. Diakinesis showing ring formation. X 13 50. 

Fig. 6. Coix gigantea. Diakinesis showing satellite bivalent. X 1800. 

Fig. 7. Coix gigantea. Anaphase I. X 13 50. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 12 




















% * 



• •• 











Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 13 


■ : 








■■ ■ ** ■'-•'■ " "-*-: i-:-: 1 

■ .-. -- ■ ■ ■ . — ■ 


[Vol. 42, 1955] 


Explanation of Plate 


Fig. 1. Coix gigantea Koen. ex Roxb. Prophase II showing extra bodies. X 990 

Fig. 2. Coix gigantea. Anaphase II. X 1170. 

Fig. 3. Polytoca macrophylla Benth. Diplotene. X 875. 

Fig. 4. Polytoca macro p by lla. Diakinesis. X 13 50. 

Fig. 5. Polytoca macrophylla. Telophase I. X 1100. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 14 


^U M^o a w »*< M^ 

4 . . 








[Vol. 42, 1955] 


Explanation of Plate 


Fig. 1. Poly toca macrophylla Benth. Prophase II. X 1170. 

Fig. 2. Poly toca macrophylla. Telophase II. X 800. 

Fig. 3. Chionachne Koenigii (Spreng.) Thwaites. Diakinesis. 

Fig. 4. Chionachne Koenigii. Anaphase I. 

Fig. 5. Chionachne Koenigii. Telophase L 

Fig. 6. Chionachne Koenigii. Anaphase II. X 15 50. 

Fig. 7. Chionachne Koenigii. Telophase II. X 1350. 




In 1947-1948, the Australian National Antarctic Research Expedition estab- 
lished weather stations at Heard Island and Macquarie Island and somewhat later 
at Mawson in MacRobertson Land, on the Antarctic Continent. Through the 
kindness of Dr. H. W. Jessep of the National Herbarium, Melbourne Botanic 
Garden, and Dr. A. M. Gwynn, Medical Officer and Biologist of the Australian 
National Antarctic Research Expedition (A.N.A.R.E.), we have had the privilege 
of studying the lichen collections. 

L Additions to the Lichen Flora of Heard Island 

Previous collections were reported by Dodge (1948) based on the British 
Australian New Zealand Antarctic Research Expedition (BA.N.Z.A.R.E.), No- 
vember 27-December 2, 1929, all between Atlas Cove and Corinthian Bay. The 
weather station was established at Atlas Cove in December 1947 (Scholes, 1951) 
and closed in 1955. The earlier collections received in 1949 were by D. Alan 
Gilchrist, Medical Officer; the collector of the later specimens was not recorded 
on the herbarium labels and are cited: A.N A.R.E. The island was more accurately 
mapped in 1948 (A.N.A.R.E. 1949). Most of the southern part of the island 
is covered by glaciers and ice fields, so that the lichen collections have come from 
the northwestern part, especially the Cape Laurens peninsula on the northwest 
corner. Thirty- two species are represented, of which three are new and fifteen 
have not been previously reported although known from Kerguelen Island to the 
northwest, making a total of 52 species known from Heard Island. 

Thelidium heardense Dodge, B.A.N.Z.A.R.E. Rept. B. 7:44. 1948. 
The thallus is lighter (vinaceous buff) than the type, but it agrees micro- 

North of Cape Laurens on volcanic rock, A.N.A.R.E. 75. 

Thelidium praevalescens (Nyl.) Zahlbr., Deutsche Sudpolar Exp. 8:51. 

Verrucaria praevalescens Nyl., in Crombie, Jour. Linn. Soc. Bot. 15:192. 1876. 
As in most previous collections, all of our material is sterile although the 

thallus has a characteristic appearance. 

West Bay, A.N.A.R.E. 743; north of Cape Laurens, on broken lava, in cave, 

A.N.A.R.E. 67, 72. 

* Graduate Student, Henry Shaw School of Botany of Washington University. 

Issued June 23, 1955. 



[Vol. 42 

Microglaena kerguelena (Nyl.) Zahlbr., Deutsche Siidpolar Exp. 8:51. 



, Bot. Brit. For. 14:22. 1876. 
Wertbii Zahlbr. New to Heard 


Xanthopyrenia heardensis Dodge & Rudolph, n. sp. 

Type: Heard Island, north of Cape Laurens at base of black cliffs, A.N.A.R.E. 


Thallus areolatus, areolis 0.3-0.5 mm. diametro subconvexis substipitatisque, 

marginibus liberis crenulatis, humectatis obscure flavo-viridibus, siccatis flavo- 

citrinis, homoeomericus; algae Xantbocapsa, coloniis sphaericis aut oblongis, vaginis 

flavo-brunneis, cellulis 7—8 xx diametro; hyphae 2-3 /x diametro inter colonias 

Perithecia immersa, 1—3 in quaque areola, subsphaerica, 200—250 xx diametro, 
ostiolo minuto; parathecium obscure brunneum, 15-20 /x crassitudine, cellulis 
polyhedricis; asci 12-15 X H5 /x, leptodermei; ascosporae octonae, imbricatim 
monostichae, late fusiformes, 19-24 X 8-9 tx, tenui cum halone dum in asco, dein 
liberae 24 X 10-11 /x, biloculares, cellula superiori majori, hyalinae, septo con- 

Thallus areolate, areoles 0.3-0.5 mm. in diameter, slightly convex above, sub- 
stipitate below, margins free, crenulate, dark yellow-green when moist, old gold 
to buffy citrine when dry, homoeomerous ; algae Xantbocapsa, colonies rounded to 
oblong, of 4-16 cells with a thick yellow-brown sheath at first, becoming densely 
packed in a homogeneous gel with abundant hyphae and cells more rounded, 7-8 /x 
in diameter, each with its own sheath about 2 /x thick; hyphae 2-3 /x in diameter, 
filling most of the interstices between the algal cells and colonies. 

Perithecia immersed or nearly so, 1-3 per areole, showing as minute dark brown 
to black points; subspherical, about 200-250 /x in diameter, ostiole small; wall dark 
brown, 15-20 /x thick, of polyhedral cells; asci 115 X 12-15 /x, thin-walled, 8- 
spored; ascospores imbricately monostichous, broad fusiform, 19-24 X 8-9 /x, 
with a thin halo while still in the ascus, 24 X 10-12 /x when free, bilocular, the 
upper cell larger, hyaline, slightly constricted at the septum. 

Steinera glaucelxa (Tuck.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:66. 1948. 

Pannaria glaucella Tuck., Bull. Torrey Bot. Club 6:57. 1875. 

Growing with Placopsis bicolor (Tuck.) B. de Lesd., on broken lava, north of 
Cape Laurens, A.N.A.R.E. 8o. New to Heard Island. 

Steinera Werthii Zahlbr., Deutsche Siidpolar Exp. 8:43. 1906. 
D. Alan Gilchrist 5 and unnumbered specimen. 

Lichina Antarctica Crombie, Jour. Bot. Brit. For. 14:21. 1876. 

On crystalline rock (sandstone?), A.N.A.R.E. New to Heard Island. 



? Siphulastrum cladinoides Dodge, B.A.N.Z.A.R.E. Rept. B. 7:69. 1948. 

"We have doubtfully referred our material to this species. It forms compact 
hemispheric cushions up to 2.5 cm. in diameter. The habit resembles young dense 
tufts of Sphaerophorus fragilis Pers. from the Arctic and Subarctic but it is com- 
pletely different in structure. The thalli are stiffer than in the type. Although 
the tips of the ultimate branches become very dark brown, we have been unable 
to find any reproductive structures. It may represent a new species. 

North of Cape Laurens on broken lava, A.N.A.R.E. 78. New to Heard Island. 

PI. 15, fig. 3. 

Collemopsidium pyrenuloides Dodge & Rudolph, n. sp. 

Type: Heard Island, north of Cape Laurens, on humus in cave, A.N.A.R.E. 65. 

Thallus gelifactus, flavidus, siccitate fragillimus, membrano-foliosus, semi- 
pellucidus, lobis irregularibus sinibus rotundatis, erectis, subrugosis; homoeomericus; 
algae Xanthocapsoideae, cellulis 5-6 ^ diametro; hyphae sparsae, tenues. 

Apothecia lecanorina immersa, 0.4-0.5 mm. diametro; amphithecium non bene 
evolutum; hypothecium ca. 25 ^ crassitudine, hyphis periclinalibus intertextis; 
thecium 100 ^ altitudine; paraphyses 3-4 /* diametro, pachydermeae; asci clavati, 
juventute apice incrassati; ascosporae octonae, brunneae, late fusiformes, biloculares, 
septo constrictae, pachydermeae, 25-30 X 10-14 /a. 

Thallus a yellowish gel, very fragile when dry, yellow, foliose-membranous, 
semipellucid, lobes very irregular with somewhat rounded sinuses, erect, subrugose; 
homoeomerous ; algae Xanthocapsoid, cells mostly singly dispersed in the gel with- 
out evident sheath, 5-6 ^ in diameter, occasionally in small subsphencal colonies 
up to 40 M in diameter with thin sheaths about each cell and a somewhat thicker 
sheath surrounding the colonies, best seen in the amphithecium; hyphae very slen- 

moss (?) fragments. 

Apothecia lecanorine, immersed or nearly so, about 0.4-0.5 mm. in diameter; 
amphithecium not clearly differentiated, a zone of Xanthocapsoid colonies with 
more abundant subvertical hyphae; hypothecium about 25 ^ thick of interwoven 
periclinal hyphae; paraphyses 3-4 M in diameter with thick walls, occasionally 
branched; thecium 100 ,* tall; asci clavate, thickened at the tip when young, dif- 
fluent, 8-spored; ascospores brown, broad, fusiform, sometimes flattened on one 
side, bilocular, constricted at the septum, wall thick, protoplasts rounded, rarely 
2 distinct protoplasts in each cell, resembling the spores of Pyrenula, 25-30 X 

ThT thallus suggests the Pyrenopsidaceae, but the algal cells mostly occur 
singly without an evident sheath and the gel is much softer when moist so that 
we have been unable to secure good sections. It is possible that the fungus is a 
parasitic Tichothecium, but the mycelium below the hypothecium is continuous 
with that of the thallus and it lacks a parathecium. Although C. Pyrenulodes is 
foliose, it seems more closely related to Collemopsidium than to any other genus 
of the Pyrenopsidaceae* 

[Vol. 42 


North of Cape Laurens, on humus in cave, A.N.A.R.E. 64, type, 67 sterile. 

Pannaria dichroa (Hook. f. & Tayl.) Crombie, Jour. Linn. Soc. Bot. 16:220, 

Lecanora dichroa Hook. f. & Tayl., London Jour. Bot. 3:643. 1844. 

Probably owing to the very rough surface of the lava, the lobes are somewhat 
ascendant and imbricate. The thallus is darker, light brownish olive and not 
stained with iron. On another very dense rock without locality, nearly covered 
by Bias tenia keroplasta Zahlbr., the thallus is quite typical. 

D. Alan Gilchrist 3; north of Cape Laurens on broken lava, A.N.A.R.E. 74' 

Lecidea Auberti B. de Lesd., Ann. Crypt. Exot. 4:99. 1931. 
D. Alan Gilchrist 8. 

Lecidea assentiens Nyl., in Crombie, Jour. Bot. Brit. For. 13:334. 1875. 
North of Cape Laurens, on volcanic rock, A.N.A.R.E. 75. New to Heard 

Lecidea subassentiens Nyl., in Crombie, Jour. Bot. Brit. For. 14:21. 1876. 
D. Alan Gilchrist 4. 

Rhizocarpon kerguelense Dodge, B.A.N.Z.A.R.E. Rept. B. 7:116. 1948. 
Jacka Valley, on cliffs, A.N.A.R.E. 33. 

Rhizocarpon Mawsoni Dodge, B.A.N. Z.A.R.E. Rept. B. 7:115. 1948. 
D. Alan Gilchrist 7. New to Heard Island. 

Rhizocarpon Johnstoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:118. 1948. 
On moraine of Schmidt Glacier, A.N.A.R.E. 242. New to Heard Island. 

Cladonia phyllophora (Tayl.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:132. 

Cenomyce phyllophora Tayl. in Hook. f. & Tayl., London Jour. Bot. 3:652. 

North of Cape Laurens, A.N.A.R.E. 70. 

Cladonia Johnstoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:135. 1948. 

Some podetia are nearly clothed with coarse granules, rarely almost isidioid, thus 
somewhat resembling C. pyxidata (L.) Fr. 

Among mosses on broken lava, north of Cape Laurens, A.N.A.R.E. 6q. New 
to Heard Island. 

Argopsis cymosa (Crombie) Stzbgr., Ber. Th'atigk. St. Gall. Naturw. Ges. 
1889-90:231. 1891. 

Stereocaulon cymosum Crombie, Jour. Linn. Soc. Bot. 15:182. 1876. 

On broken lava, north of Cape Laurens, A.N.A.R.E. 71. New to Heard Island. 

Ureceolina kergueliensis Tuck., Bull. Torrey Bot. Club 6:58. 1875 
Jacka Valley, 600 ft 

— ^ -- — — — — -j — - - — — - _____ ___ 

, on cliff, A.N.A.R.E. 34. New to Heard Island. 



Aspicilia lygomma (Nyl.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:164. 1948. 

Lecidea lygomma Nyl. in Crombie, Jour. Bot. Brit. For. 13:3 34. 1875. 

On boulders, 20 ft., West Bay, A.N.A.R.E. 743, growing with Thelidium 
praevalescens (Nyl.) Zahlbr. and Kuttlingeria crozetica (Zahlbr.) Dodge. New 
to Heard Island. 

Aspicilia disjunguenda (Nyl.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:167. 

Lecanora disjunguenda Nyl. in Crombie, Jour. Bot. Brit. For. 15:105. 1877. 
D. Alan Gilchrist 6, 7, p. 


Aspiciliopsis macrophthalma (Hook. f. & Tayl.) Dodge, B.A.N.Z.A.R.E. 
Rept. B. 7:175. 1948. 

Urceolaria macrophthalma Hook. f. & Tayl., London Jour. Bot. 3:640. 1844. 

D. Alan Gilchrist 10. 

Placopsis bicolor (Tuck.) B. deLesd. Ann. Crypt. Exot. 4:100. 1931. 

Placodium bicolor Tuck., Bull. Torrey Bot. Club 6:57. 1875. 

The specimens are much paler than usual, probably from less iron in the rocks, 

and cephalodia are very rare. 

North of Cape Laurens, on broken lava, A.N.A.R.E. 8o, growing with Strinera 
glaucella (Tuck.) Dodge; on cliff, Jacka Valley, 600 ft., A.N.A.R.E. 31. 

Usnea Taylori Hook, f., London Jour. Bot. 3:657. 1844. 

Moraine of Baudessen Glacier, 800 ft., A.N.A.R.E. 250. New to Heard Island. 

Usnea insularis (Lamb) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:211. 1948. 

Neuropogon insularis Lamb, Jour. Linn. Soc. Bot. 52:215. pi. 8, fig. 17. 1939. 

Only a few plants, beginning to form cupulate apothecia. 

Mt. Aubert de la Riie, 300 ft., A.N.A.R.E. 4. 

Usnea trachycarpa (Stirton) Mull. Arg. Nuovo Giorn. Bot. Ital. 21:37. 



Neuropogon trachycarpus Stirton, Scottish Nat. 6:105. 
D. Alan Gilchrist 2. New to Heard Island. 

Blastenia keroplasta Zahlbr., Deutsche Siidpolar Exp. 8:28. 1906. 
Two specimens without locality, A.N.A.R.E. New to Heard Island. 





A.N.A.R.E. 743; on cliff in Jac 
christ I. New to Heard Island. 

bplicata Nyl. in Crombie, J 

Jahrb. [Engler] 5:138. 1884. 
Rnt. For. 13:334. 1875. 


[Vol. 42 


Rinodin a aspicilin a Zahlbr., Deutsche Siidpolar Exp. 8:50. 1906. 
Two small specimens without locality, A.N.A.R.E. 


Occasionally lichenologists have encountered conidial fructifications on lichen 
thalli, sometimes associated with apothecia, sometimes not. Miiller-Argau (1881) 
described an otiform conidial structure which he named a campylidium. Vainio 

found the same structure on a thallus of Lopadium perpallida (Nyl.) 


pallida Nyl.) and referred it to Cyphella 

Spegazzini (1909) 




Karst. and described several new varieties. Mameli Calvino (1930) proposed the 
name Deuterolichenes to include Cblorocyphella, and Cengia Sambo (1937, 1941) 
and Rizzini (1952) have reported species of Cblorocyphella not associated with 
apothecia. Malme (1935) considered C. aeruginascens (Karst.) Keissl. to be a 
conidial stage of Lopadium per pallidum (Nyl.) Zahlbr., and Dodge (1953) de- 
scribed a campylidium as a conidial state of L. Deightoni Dodge. Campylidia have 
also been observed on Sporopodium sp. 

Miiller-Argau (1890) described another type of fructification, the orthidium, 
which resembles an apothecium except that the thecium is replaced by conidiophores, 
and the senior author has seen a similar fructification on a foliicolous thallus from 
Jamaica. Our material contains an orthidium-bearing lichen from Heard Island, 
resembling Ephelis Fr. of the Excipulaceae, which has unicellular spores, while those 
of our lichen become septate as do those of Ephelis trinitensis Cooke & Massee, the 
imperfect state of Balansia trinitensis Cooke & Massee on Panicum palmifolium in 


EPHELIDIUM Dodge & Rudolph, n. gen. 


Thallus crustosus, indeterminatus, sorediosus, ecorticatus, heteromericus ; algae 
protococcoideae. Orthidium sessile, concavum, margine persistente; conidiophorae 
simplices; conidia singulatim disposita, terminalia, acicularia, septata. 

Thallus crustose, indeterminate, sorediose, ecorticate, heteromerous; algae 
protococcoid. Orthidium sessile, concave with a persistent margin, resembling a 
lecanorine (or biatorine as the algae die above) apothecium; conidiophores un- 
branched; conidia single, terminal, acicular, long remaining unicellular but finally 

Ephelidium heardense Dodge & Rudolph, n. sp. Pi. 15, fig- !• 

Type: Heard Island, Atlas cove at foot of Poa mound, A.N.A.R.E. 147. 

Thallus crustosus, indeterminatus, 0.7-1.2 mm. crassitudine, citrinus, sore- 
diosus; ecorticatus; stratum algarum ca. 280 ft crassitudine, cellulis protococcoideis, 
8.7-12.2 /x diametro; medulla crassa, hyphis 1 /x diametro dense intertextis, nubilatis. 



Orthidium sessile, basi constrictum, orbiculare, 0.6—1.5 mm. diametro, sub- 
ochraceum, margine 250-280 /x crassitudine; conidiophorae tenues, unicellulares, 
13—17 fi longitudine; conidia singulatim disposita, acicularia, hyalina, 30—44 X 2 /x, 
primum unicellulares, dende ad 7-septata, recta aut subcurvata. 

Thallus crustose, indeterminate, 0.7-1.2 mm. thick, buffy citrine, sorediose, 
K orange brown, C — ; ecorticate; algal layer about 280 /x thick, cells protococcoid, 
spherical to somewhat polyhedral from mutual pressure, 8.7-12.2 /x in diameter; 
medulla thick, of closely woven hyphae about 1 /x in diameter, somewhat nubilated 
with granules and including pieces of roots etc. from the substrate. 

Orthidium sessile, constricted at the base, circular, 0.6-1.5 mm. in diameter, 
ochraceous buff, disc very concave; margin 250-280 /x thick, finally undulate; 
algae in a discontinuous layer on the outside of the layer of conidiophores, tending 
to die out above and forming a continuous layer below; the medulla around and 
between the algal colonies is formed of compactly woven hyphae about 1 /x in 
diameter; conidiophores arising from the medullary hyphae, forming a continuous 
layer 13-17 tt thick; conidia borne singly, acicular, ; 

septate until late, finally up to 7-septate, straight or slightly curved, slightly taper- 
ing at the ends. 


II. Additions to the Lichen Flora of Macquarie Island 

Previous collections were reported by Dodge (1948). Most of the present 
collections are from the northern half of the island, collected mostly by Norman 
R. Laird and by N. M. Haysom of the A.N.A.R.E. Taylor (1954) has discussed 
the problem of distribution of the flowering plants which apparently have much 
wider ranges than the lichens. Twenty-seven species are reported from the present 
collection, three of which are new and three not previously reported bringing the 
total species of lichens to forty-four. Several other possible new species will be 
reported in a later number of these Lichenological Notes. In the following list 
MI/49/ has been omitted from N. M. Haysom's numbers of collections. 

Microthelia macquariensis Dodge, B.A.N. Z.A.R.E. Rept. B. 7:48. 1948. 



Orange red when fresh, drying olive buff. The Trentepohlia filaments have 
few corticating hyphae. From glacial moraine above Sandy Bay, 500 ft., 

very few corticating hyphae. 
N. M. Haysom Z2. 

Psoroma versicolor (Hook. f. & Tayl.) Mull. Arg., Flora 71:53 8. 1888. 
Lecanora versicolor Hook. f. & Tayl., London Jour. Bot. 3:642. 1844, non Ach. 
The squamules are less well developed than in previous collections but the 

apothecia agree microscopically. 

Norman R. Laird 2, 2a, 2c; Featherbed Terrace, growing over decaying hepatics 

and other vegetable debris, A.N.A.R.E. 

[Vol. 42 


Pseudocyphellaria glabra (Hook. f. & Tayl.) Dodge, B.A.N.Z.A.R.E. Rept. 
B. 7:79. 1948. 

Stricta glabra Hook, f, & Tayl., London Jour. Bot. 3:647. 1844. 

South end of Plateau, on moss, N. M. Haysom Z114; north end of Plateau, on 
soil and over decaying mosses, Norman R. Laird; Featherbed Terrace, A.N.A.R.E.; 
Wireless Hill, N. M. Haysom Zi; north of Lusitania Bay, N. M. Haysom Z134. 

Peltigera Lairdi Dodge & Rudolph, n. sp. 

Type: Macquarie Island, growing on soil over decaying grasses, mosses and 
hepatics, Norman R. Laird 4, A.N.A.R.E. 

Thallus foliosus, in herbario isabellinus, ad 10 cm. diametro, 400 /x crassitudine, 
lobis sterilibus ca. 1.5 cm. latitudine, 2 cm. longitudine; tomentum crassum, verru- 
cosum in partibus junioribus; superficies inferior arachnoideus inter venas; venae 
sparsae, elevatae, arachnoideo-tomentosae, rhizinis simplicibus vel semel dichotome 
ramosis, ca. 5 mm. longitudine, concoloribus; cortex ca. 90 /x crassitudine, pseudo- 
parenchymaticus, cellulis 12—15 /x diametro leptodermeis; stratum algarum 50—55 /x 
crassitudine, cellulis nostocaceis, 5 /x diametro; medulla 270 it crassitudine, hyphis 
periclinalibus pachydermeis, 4 xt diametro, parte superiore compacta, inferiore laxi- 
ore, hyphis ad 8 ft diametro, lumine 5 /x. 

Apothecia in lobis erectis non revolutis, 1 5 mm. altitudine, 5-6 mm. latitudine, 
sita; discus nigro-brunneus, 4 mm. diametro; sine amphithecio nee parathecio, 
stratum algarum sub hypothecio crassiore, ad 120 /x crassitudine; hypothecium 
30 /x crassitudine, brunneum, superne obscurius, pseudoparenchymaticum, hyphis 
periclinalibus; thecium 120-130 ju altitudine; paraphyses simplices, ca. 2 /x diametro; 
asci 100-105 X 11—12 /x, juventute apice incrassati; ascosporae octonae, hyalinae, 
aciculares, multiseptatae, ca. 56 X 3 /x. 

Thallus foliose, becoming isabella color in the herbarium, up to 10 cm. in 
diameter, 400 /x thick, margin lobed, sterile lobes about 1.5 cm. wide and 2 cm. 
long, somewhat lacerate on drying, ends of lobes with thick, verrucose tomentum 
above, becoming smooth but surface dull in the older portions as the tomentum 
weathers away; underside pinkish buff, arachnoid between the sparse veins; veins 
elevated, arachnoid- tomentose; rhizinae not abundant, simple or once-dichotomous 
near the tips, about 5 mm. long, concolorous; cortex about 90 /x thick, pseudo- 
parenchymatous, cells rather thin-walled, 12-15 /x in diameter; algal layer 50-55 fi 
thick, of N os toe colonies, cells about 5 xt in diameter in an inconspicuous gel; 
medulla 270 /x thick, of closely packed periclinal hyphae, thick-walled, about 4 /x 
in diameter, less compact and more irregularly arranged below, forming the arach- 
noid underside with hyphae up to 8 xx in diameter, lumen 5 /x. 

Apothecia on erect lobes, not revolute, 15 mm. tall, 5-6 mm. wide, disc very 
dark brown, 4 mm. in diameter; no amphi thecium nor para thecium differentiated; 
algal layer thicker below the hypothecium, up to 120 /x thick; hypothecium 30 /a 
thick, lower half pale brown, upper half very dark brown, pseudoparenchymatous 
from periclinal hyphae; thecium 120—130 tx tall; paraphyses about 2/x in diameter, 
unbranched, tips ending in the dark brown epithecial gel; asci 100-105 X 11—12 /*> 



tips slightly thickened when young, 8-spored; ascospores hyaline, acicular, 56 X 
3 /x, thin-walled, many septate. 

This species somewhat resembles Peltigera dilacerata Gyelnik from Auckland, 
New Zealand, which has longer and narrower lobes with dilacerate margins and is 
about half as thick. 

Lecidea subglobulata Knight, Trans. Proc. N. Zeal. Inst. 8:314. 1875 

South slopes of Wireless Hill, A.N.A.R.E. New to Macquarie Island. 

Catillaria (Eucatillaria) Rudolphi Dodge, n. sp. 

Type: Macquarie Island, without locality, but the rock types closely resemble 
those from the south slopes of Wireless Hill, A.N.A.R.E. 

Thallus crustosus, albidus, continuus aut ad centrum areolatus crassior, mar- 
gine irregulariter lobatus; cortex fastigiatus subdecompositus, erosus, ad 30 /x 
crassitudine; stratum algarum 65 /x crassitudine, cellulis protococcoideis 6-8 xt 
diametro; medulla 250-300 /x crassitudine, hyphis tenuibus, granulis brunneis 

nubilatis, dense contexta. 

Apothecia subimmersa aut sessilia, ad 2.5 mm. diametro, orbicularia, aut mutua 
pressione angularia, margine elevato, disco piano aut subconvexo nigro; para- 
thecium carbonaceum, in margine 50 /x crassitudine ad 125-160 /x sub thecio; 
hypothecium non bene evolutum; thecium 90-120 /x altitudine; paraphyses tenues, 
septatae, super ascos ramosae, apicibus clavatis brunneis, ca. 1.5 tt diametro; asci 
50-60 X 7-8.5 /x, cylindrico-clavati, leptodermei; ascosporae octonae, hyalinae, 
biloculares, uniseriales, ellipsoideae, 13—15 X 6—7 /x. 

Thallus whitish, sometimes stained ferruginous from iron in the underlying 
rock, thin, continuous at the margin, thicker toward the areolate center, K yellow 
then orange, margin irregularly lobate, distinct; cortex fastigiate, somewhat de- 
composed and eroded, up to about 30 /x thick; algal layer 65 /x thick, cells pro to- 
coccoid, 6-8 /x in diameter in a continuous dense layer; medulla 250-300 xi thick, 
of densely woven slender hyphae, heavily nubilated with dark brown granules. 

Apothecia slightly immersed to sessile, up to 2.5 mm. in diameter, abundant, 
circular or angular from mutual pressure, margins elevated, disc plane to slightly 
convex, black; parathecium carbonaceous, 50 /x thick at the margin, 125-160 /x 
thick below the thecium; hypothecium not clearly differentiated; thecium 90-120 it 
tall; paraphyses slender, septate, branched at the level of the tips of the asci, tips 
slender, clavate, brownish, about 1.5 /x in diameter, asci 50-60 X 7-8.5 it, cyiindric- 
clavate, thin-walled, 8-spored; ascospores hyaline, bilocular, mostly uniseriate, 

ellipsoidal, 13-15 X 6-7 /i. 

The germinating ascospore produces an extensive black hypothallus. As con- 
tact is made with algal cells, thin assimilative areoles develop and finally fuse, 
covering the hypothallus as a continuous thallus which gradually thickens with age. 

Without locality but probably from south slopes of Wireless Hill, A.N.A.R.E., 
type; west side of Wireless Hill, N. M. Haysom Z152; south end of Plateau, N. M. 


[Vol. 42 

Haysom Z137; Plateau, 900 ft., young, A.N.A.R.E.; north end of Plateau, N. M. 
Haysom Z86; without locality, Norman R. Laird 8. 

Cladia aggregata (Sw.) NyL, Bull. Soc. Linn. Normand. II, 4:167. 1870. 

Lichen aggregatus Sw., Nova Gen. Sp. Pi. Ind. Occ. 147. 1788. 
West coast, fluviatile swamp, 50 ft., A.N.A.R.E. 

Cladonia floriformis Dodge, B.A.N.Z.A.R.E. Rept. B. 7:134. 1948. 

Wireless Hill, 300 ft., Norman R. Laird id; Nuggets Creek, Norman R. Laird 
la, lb, young; without locality, Norman R. Laird I; only a few young plants 
among mosses, A.N.A.R.E. $c. 



Nuggets Valley, 100 ft., A.N.A.R.E.; ? primary thallus only, mouth of cave, 
south end of isthmus, A.N.A.R.E. 

Cladonia sarmentosa (Tayl.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:129. 1948. 

Cenomyce sarmentosa Tayl. in Hook. f. & Tayl., London Jour. Bot. 3:651. 

Interior of cave, 35 ft., A.N.A.R.E.; near Brothert Point, 250 ft., A.N.A.R.E.; 
near Nuggets Creek, Norman R. Laird ic, id, 2c; Lusitania Valley, east coast, 
150 ft., A.N.A.R.E. 

Stereocaulon argodes NyL, Compt. Rend. Acad. Sci. [Paris] 83:88. 1876. 

Stereocaulon Argus Th. Fr., Nova Acta R. Soc. Sci. Upsal. Ill, 2 1 :333. 1858; 
Hook. f. & Tayl., London Jour. Bot. 3:653. 1844, pro parte min. 

For discussion of nomenclature, see Dodge (1948, pp. 142-144) sub Argopsis 
megalospora. In the present collection the dimensions of the ascospores are some- 
what greater, 45-53 X 5.5-7 /x, than those given by Th. M. Fries and Nylander. 

On coastal vegetation north of Lusitania Bay, N. M. Haysom Z128; Plateau, 
N. M. Haysom Zi. 

Stereocaulon pulvinare Dodge, B.A.N.Z.A.R.E. Rept. B. 7:139. 1948. 
Near Nuggets Creek, Norman R. Laird. 


Stereocaulon submollescens NyL, Compt. Rend. Acad. Sci. [Paris] 83:88. 

Featherbed Terrace, A.N.A.R.E.; north end of Plateau, N. M. Haysom Z85. 

Pertusaria tyloplaca Nyl. Compt. Rend. Acad. Sci. [Paris] 83:90. 1876. 

South end of Plateau, N. M. Haysom Z122; Plateau, 900 ft., A.N.A.R.E.; 
north end of Plateau, N. M. Haysom Z59, Z62; growing over Menegazzia circum- 
sorediata Santesson on radio mast erected by A.A.E. in 1911 on Wireless Hill, 

Placopsis perrugosa (Nyl.) NyL, Lich. Nov. Zelandiae, 57. 1888. 



For complete description, see I. M. Lamb, Lilloa 13:268-272. 1947. 



On rocky banks of Lusitania Creek, N. M. Haysom Z123. New to Macquarie 



Type: Macquarie Island, Wireless Hill, on cliff subject to water seepage, 

Thallus fruticosus erectus aut subdecumbens, pulvinos hemisphaericos ad 3 cm. 
diametro formans, di- aut tri-chotome ramosus, torulosus, cinnamomeo-alutaceus, 
inferne ad 0.8 mm. diametro, superne tenuior, ad 1 cm. altitudine; ecorticatus; 
algae flavo-virides, cellulis sphaericis vel mutua pressione polyhedricis, 8-11 p. 
diametro; hyphae medullares ca. 3 fi diametro, conglutinatae, inter cellulas algarum. 

Apothecia ad 1 mm. diametro in lateribus ramorum sessilia, plana dein convexa 
emarginataque, ca. 250 fi altitudine, disco nigro; amphithecium 250 fi crassitudine; 
parathecium non bene distinctum, margine ad 40 /* crassitudine; hypothecium 
obconicum, ca. 190 fx. altitudine, hyphis verticalibus nigro- brunneis; thecium 135 /t 
altitudine; paraphyses 2.5-3 /a diametro, septatae, cellula terminali clavata 4 X 
5.5 ju., pachydermea nigro-brunnea ; asci ca. 65 X 12 /*, clavati, juventute apice 
incrassati; ascosporae octonae, subdistichae in ascis 16 X 8 /*> biloculares, brunneae, 
septo tenui constrictae, liberatae, 12 X 6 /*, nigro-brunneae. 

Thallus fruticose, erect or subdecumbent, forming dense, depressed hemispheric 
cushions up to 3 cm. in diameter, branching di- or trichotomous, torulose, pinkish 
buff to cinnamon buff, up to 0.8 mm. in diameter below, more slender above, 
circular in cross-section, about 1 cm. tall; ecorticate; algae yellow-green, spherical 
or somewhat polyhedric from mutual pressure in the outer portion, 8-11 /1 in 
diameter with a thick sheath of periclinal medullary hyphae, more scattered 
throughout the medulla; hyphae about 3 fi in diameter, compact and conglutinate 

between the algal cells. 

Apothecia up to 1 mm. in diameter, sessile on the sides of the branches, not on 
the ultimate branches, flat at first becoming convex and emarginate, about 250 /* 
tall, disc black; amphithecium about 250 fi thick, similar in structure to that of 
the thallus but the medullary hyphae more vertical; parathecium not well differ- 
entiated, about 40 fi thick at the margin, scarcely differentiated from the para- 
physes, continuous below with the dark brown obconic hypothecium about 190 » 
tall at the center, of vertical dark brown hyphae, not sharply differentiated from 
the thecium above; thecium 135 /1 tall; paraphyses 2.5-3 fi in diameter, septate, 
terminal cell broadly clavate, about 4 X 5.5 /*, thick-walled, dark brown above; 
asci about 65 X 12 /*, clavate, tips very thick when young, 8-spored; ascospores 
subdistichous, 16 X 8 /t, bilocular, brown, ellipsoid, slightly constricted at the 
thin septum, shrinking to 1 2 X * A* when free from the ascus and very dark brown. 

The specimens have been cut from the underlying rock, but the cut ends sug- 
gest that the base may expand into a circular holdfast. The systematic position of 
this species is not clear. The bilocular brown spores with a thin septum, the para- 
physes and lecanorine apothecium suggest Rinodina sect. Beltrammta; the very 
dark hvoothecium suggests Melants pi cilia Vainio. From both of these it differs in 


[Vol. 42 

its fruticose thallus. The structures of the thallus and of the apothecia are wholly 
different from those of the fruticose species of Anaptychia, but do resemble those 
of T hamnolecania from which it differs in a very dark brown hypothecium and 
broader brown bilocular spores. Its yellowish green algae suggest a relationship 
to Thelidea Hue, which is foliose with biatorine apothecia and hyaline ascospores. 
It is not a parasite as the medullary hyphae are continuous with those at the base 
of the hypothecium and the amphithecium is well developed; hence it cannot be 
considered a Karschia sp. on a Sipbula. 

Parmelia Turgidula Bitter, Hedwigia 40:246. 1901. 

Featherbed Terrace, over mosses. A single sterile plant. New to Macquarie 
Island but previously known from New Zealand. 

Parmelia sublugubris Dodge, B.A.N.Z.A.R.E. Rept. B. 7:188. 1948. 
Featherbed Terrace, A.N.A.R.E.; Norman R. Laird 3; north end of Plateau, 
N. M. Haysom Z.61. 

Parmelia tenuirima Hook. f. & Tayl, London Jour. Bot. 3:645. 1844. 
Featherbed fluviatile terrace on northwest slopes, growing over mosses, Norman 
R. Laird; from rock face, Half Moon Bay, west coast, A.N.A.R.E. 

Menegazzia circumsorediata Santesson, Ark. f. Bot. 30:11:14. 1942. 
Thallus 9 cm. in diameter. Wireless Hill, on radio mast erected by the A.A.E. 
in 1911, A.N.A.R.E.; coastal rocks north of Lusitania Bay, N. M. Haysom Z127. 

Usnea arida v. muscicola Dodge, B.A.N.Z.A.R.E. Rept. B. 7:207. 1948. 
Wireless Hill, on planks of A.A.E. radio hut, N. M. Haysom Z53. 

Usnea contexta Motyka, Lich. Gen. Usnea Stud. Monog. 2:436. 1937. 
Featherbed Terrace, A.N.A.R.E.; north end of Plateau, N. M. Haysom Z83. 

Usnea torulosa (Mull. Arg.) Zahlbr., Cat. Lich. Univ. 6:594. 1930. 




North of Lusi- 


Usnea xanthopoga Nyl., Compt. Rend. Acad. Sci. [Paris] 83:89. 1876. 
Among mosses, Norman R. Laird 5a, 5b, 5c, $d t $e; Featherbed Terrace, very 
young, A.N.A.R.E. 

Ramalina geniculata Hook. f. & Tayl., London Jour. Bot. 3:665. 1844. 

Turf from rock surface, west coast, 50 ft., A.N.A.R.E.; Featherbed Terrace, 

Gasparrinia macquariensis Dodge, B.A.N.Z.A.R.E. Rept. B. 7:234. 1948. 

One plant has much larger apothecia, up to 4 mm. in diameter, with flat discs 
and nearly excluded margins but agrees microscopically. 

Hasselborough Bay, isthmus beach, Norman R. Laird 8a; beach of Garden Cove, 
N. M. Haysom Zn, Z223; Green Gorge, N. M. Haysom Z102. 



Buellia Mawsoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:243. 1948. 
South end of Plateau, N. M. Haysom Z/J7. 

III. Flora of Adelie Land 

A small collection was received from Cape Margerie, 66° 50' S., 141° 20' E., 
collected on November 4, 1950, the first specimens taken from Adelie Land. Since 
the collection is small and Adelie Land lies between King George V Land and Queen 
Mary Land, it is not surprising that all these species have been reported before. 

Umbilicaria Hunteri Dodge, B.A.N.Z.A.R.E. Rept. B. 7:148. 1948. 
A.N.A.R.E. 4. Previously known from King George Land. 

Alectoria congesta (Zahlbr.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:195. 1948. 

Parmelia pubescens v. congesta Zahlbr., Deutsche Siidpolar Exp. 8:52. 1906. 
A.N.A.R.E. J. Previously known from King George V Land to MacRobertson 

Usnea Antarctica DuRietz, Svensk Bot. Tidskr. 20:90, 93. 1926. 
A.N.A.R.E. 6. Previously known from Marie Byrd Land to King George V 

Usnea pustulata Dodge, B.A.N.Z.A.R.E. Rept. B. 7:203. 1948. 
A.N.A.R.E. 5a. Previously known from King George V Land and Queen 

Mary Land. 

Usnea scabridula (Lamb) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:204. 1948. 

Neuropogon acromelanus v. inactivus f. scabridulus Lamb, Jour. Linn. Soc. Bot. 
52:220. 1939. 

A.N.A.R.E. 5. Previously known from South Victoria Land to King George V 


Xanthoria Mawsoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:236. 1948. 
A.N.A.R.E. I. Previously known from King George V Land to MacRobertson 


Rinodina frigida (Darb.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:259. 1948. 
Buellia frigida Darb., Brit. Nat. Antarct. "Discovery" Exp. Nat. Hist. 5:Lich.:7. 


A.N.A.R.E. 3. Previously known from Marie Byrd Land to MacRobertson 


IV. Additions to the Flora of MacRobertson Land 

Previous collections from Cape Bruce were reported by Dodge (1948). The 
present collections are from the A.N.A.R.E. weather station at Mawson, about 
67° S. and 50° E., collected by the Medical Officer, Dr. A. M. Gwynn. Fourteen 
species are represented, one new species, six others new to MacRobertson Land, 

[Vol. 42 


making twenty-four species so far known. The Verrucariaceae are still absent 
from collections. 

Heppia Antarctica Dodge, B.A.N.Z.A.R.E. Rept. B. 7:71. 1948. 
Small sterile thalli along with several genera of Myxophyceae on a weathered 
crystalline rock with Parmelia Johnstoni Dodge, A. M. Gwynn L123. 

Lecidea Harrissoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:101. 1948. 
On rock with Alectoria conges ta (Zahlbr.) Dodge, A. M. Givynn L129. 

Umbilicaria spongiosa v. subvirginis (Frey & Lamb) Dodge, B.A.N.Z.A.R.E. 
Rept. B. 7:148. 1948. 

Umbilicaria antarctica v. subvirginis Frey & Lamb, Trans. Brit. Myc. Soc. 
22:272. 1939. 

Thallus about 7 cm. in diameter, upper surface pale smoke-gray or lighter, 
very minutely rimulose. 

A. M. Gwynn Li^o. New to MacRobertson Land. 

Umbilicaria subcerebriformis Dodge, B.A.N.Z.A.R.E. Rept. B. 7: 149. 1948. 

A. M. Gwynn Lijo. New to MacRobertson Land. 
Acarospora (Pachnolepia) Gwynni Dodge & Rudolph, n. 


PL 15, fig. 2. 

Type: MacRobertson Land, Mawson, on weathered reddish granite? A. M. 
Gwynn Li2i. 

Thallus cerebriformis subfruticulosus, juventute citrinus dein magis olivaceo- 
viridis, ad 3 mm. diametro, substipitatus, marginibus liberis, ca. 500 /1 crassitudine, 
inferne alutaceus; cortex superne 20-25 fi inferne ad 30 /x crassitudine, fastigiatus, 
pseudoparenchymaticus, cellulis pachydermeis, luminibus 2.5 jt diametro, strato 
amorpho gelifacto 15 p crassitudine tectus; algae protococcoideae, cellulis 13-14 /x 
diametro; medulla non distincte evoluta, hyphis 2 /x diametro. 

Apothecia ca. 0.5 mm. diametro, orbicularia vel mutua pressione elliptica, im- 
mersa, disco obscuro; parathecium hyalinum, 15 tt crassitudine in lateribus thecii, 
subtus 25-30 fx crassitudine, hyphis periclinalibus; hypothecium 15 it crassitudine, 
hyphis tenuibus grumosis dense contextum; thecium 230 /1 altitudine; paraphyses 
cohaerentes, distinctae, pachydermeae, 2 /i diametro; asci 100 X 20 it, juventute 
apice incrassati; ascosporae multae, ellipsoideae, ca. 3.5 X 1-8 it (liberae non visae). 

Spermogonia immersa, 200 it diametro; perifulcrum hyalinum, 15 it crassi- 
tudine, hyphis pachydermeis, periclinalibus conglutinatis; fulcrum in cavitate in- 
vaginans; spermatiophorae 12-15 it longitudine, subramosae (modo Cladoniae); 
spermatia hyalina, anguste ellipsoideae, J X 1 /*. 

Thallus cerebriform, subfruticulose, lemon-yellow in younger portions to 
warbler green when the apothecia are well developed, covering areas up to 2 cm. 
in diameter, individual thalli up to 3 mm. in diameter, attached at the center, 
margins free, about 500 it thick, upper surface smooth to deeply furrowed and 
cerebriform, under surface warm-buff, K — ; cortex 20-25 it above to 30 it thick 
below, completely surrounding the thallus, fastigiate, of thick-walled pseudo- 



parenchyma, cell lumina 2.5 li in diameter, densely nubilated with minute yellowish 
crystals, especially above, covered by an amorphous layer of gel about 1 5 fx thick, 
with scattered granules; algal layer filling the thallus between the cortices, cells 
protococcoid, 13-14 fx in diameter, often polyhedral from mutual pressure; medulla 
not differentiated, hyphae about 2 /x in diameter between the algal cells. 

Apothecia very numerous in the central thalli, about 0.5 mm. in diameter, 
circular to somewhat elliptic from mutual pressure, immersed, disc greenish black; 
amphithecium not differentiated from the thallus, but often with a circular furrow 
about 0.2 mm. from the parathecium; parathecium hyaline, of periclinal hyphae 
about 15 /Jt thick at the sides of the thecium to 25—30 fx thick below; hypothecium 
about 15 fi thick, of grumose, densely woven, slender hyphae; thecium 230 /x tail, 
the upper 25 \x brownish and covered by a hyaline amorphous gel about 13 fx thick; 
paraphyses coherent, distinct, thick- walled, about 2 fx in diameter; asci about 
100 X 20 jx> tips thickened when young; ascospores numerous, ellipsoid, about 
3.5 X 1-8 /x (not seen free from ascus). 

Spermogonia immersed, about 200 /x in diameter, wall hyaline, 15 /x thick, of 
thick-walled, conglutinate, periclinal hyphae; layer of spermatiophores invaginated, 
forming cerebriform cavities, spermatiophores 12-15 /x long, somewhat branched 
(Cladonia type) ; spermatia hyaline, narrow ellipsoid, about 3 X 1 /*• 

On weathered reddish granite, A. M. Gwynn Lt2i, Lt2y. 

Lecanora griseomarginata Dodge & Baker, Ann. Mo. Bot. Gard. 25:572. 

Growing on moss cushion, A. M . Gwynn Li20. New to MacRobertson Land. 

Lecanora exsulans (Th. Fr.) Dodge & Baker, Ann. Mo. Bot. Gard. 25:570. 

Lecanora chrysoleuca v. melanophthalma f. exsulans Th. Fr., Nyt Mag. Natur- 

vidensk. 40:208. 1902. 

A. M. Gwynn Li2l, Li24 y Li^S, U31 (a single young plant not sectioned) . 




Growing over weathered rocks and mosses, A. M. Gwynn Li2i, U22, L123, 
IJ25, Li 28, Lis I. 

Alectoria congesta (Zahlbr.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:195. 1948. 

Parmelia pubescens v. congesta Zahlbr., Deutsche Siidpolar Exp. 8:! 
On rock, A. M. Gwynn Li2Q. 

Protoplastenia citrina Dodge, B.A.N.Z.A.R.E. Rept. B. 7:222. 
The color is more orange than in the type and all the material is steril 
the thallus agrees microscopically. 

Growine over moss cushions, A. M. Gwynn Li20, Li2l, Li25. 



[Vol. 42 


Gasparrinia Harrissoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:23 5. 1948. 
Thallus about 5 cm. long and 2 cm. wide, apparently starting near the edge 
of the rock fragment, so probably capable of forming a circular thallus about 5 


cm. in diameter when the substrate permits. A. M. Gwynn U26, L131, a smaller 

Xanthoria Mawsoni Dodge, B.A.N.Z.A.R.E. Rept. B. 7:236. 1948. 




A. M. Gwynn U28, associated with Lecanora exsulans (Th. Fr.) Dodge & 
Baker, as in the type. New to MacRobertson Land. 

Rinodina frigida (Darb.) Dodge, B.A.N.Z.A.R.E. Rept. B. 7:259. 1948. 

Buellia frigida Darb., Brit. Nat. Antarct. "Discovery" Exp., Nat. Hist. 
5:Lich.:7. 1910. 

A. M. Gwynn Li2l, with lecanorine apothecia nearly immersed; L$2f $ U33, 
larger older thallus up to 6 cm. in diameter. 

Nostoc sp. 

Several thalli growing on a moss cushion, akinetes abundant. A few young 
perithecia or cleistothecia were seen which may be parasites, or perhaps the thalli 
are those of a Pyrenidiaceous lichen. Very few fungus hyphae were seen in the gel 
between the algal filaments but in one or two places there is a faint suggestion of 
a cortex. Some thalli contained a few bacterial colonies. We hope it will be found 
again in a more mature condition, if it be a lichen. 

A. M. Gwynn Lt20 y along with Lecanora griseomarginata Dodge & Baker and 
sterile Protoblastenia citrina Dodge. 


Australian National Antarctic Research Expedition (1949). [Map of] Heard Island from survey 
by Australian National Antarctic Research Expedition, 1948, scale 1:83,000. 

Cengia Sambo, Maria (1937). Lichenes Africae Orientalis Italicae. Nuovo Giorn. Bot. Ital. N.S. 
44:456-473. 3 figs, [see p. 470]. 

, (1941). Fragmenta lichenologica III. Ibid. 48:495-516. [see p. 500-501], 

Dodge, Carroll W. (1948). Lichens and lichen parasites. British Australian New Zealand Antarctic 
Research Expedition (B.A.N.Z.A.R.E.) Rept. B. 7:1-276. 

, (1953). Some lichens of tropical Africa. Ann. Mo. Bot. Gard. 40:271-401. [see 

pp. 362-363]. 

Keissler, K. v. (1927). Systematische Untersuchungen iiber die Flechtenparasiten und lichenoide 
Pilze. V. Ann. Naturhist. Mus. Wien 41:157-170. [see p. 157-163]. 

Malme, G. (1935). Das Kampylidium, ein verkanntes Fortpflanzungsorgan der Flechten. Svensk 
Bot. Tidskr. 29:302-3 05. 

Mameli-Calvino, Eva (1930). Ricerche su una forma singolare di Deuterolicheni: Chlorocyphella 
subtropica Speg. Nuovo Giorn. Bot. Ital. N.S. 37:369-379. pi. 19: 

Miiller-Argau, J. (1881). Lichenologische Beitrage. XII (Schluss). Flora 64:100-112. [seep. 111- 

— — — , (1890). Lichenologische Beitrage. XXXIII. Ibid. 73:187-202. [see p. 202]. 
Rizzini C. Toledo (1952). Lichenes in Horto Botanico Fluminis Januarii crescentes. I. Arq. Jard. 
Bot. Rio de Janeiro 12:187-202. 4 pi. 

Scholes, Arthur (1951). Fourteen men, the story of the Australian Antarctic Expedition to Heard 
Island. London front. 5 p. 1. 273 pp. 10 pi. 




Spegazzini, C. (1909). Mycetes Argentinenses. Series IV. An. Mus. Nac. Buenos Aires 19:257-458. 

40 fig. [see p. 279]. 
Taylor, B. W. (1954). An example of long distance dispersal. Ecology 35:569-572. 
Vainio, E. A. (1890). Etude sur la classification naturelle et la morphologie des lichens du Br£sil. 

[Thesis Univ. Helsingfors]. Acta Soc. Fauna Fl. Fenn. 7:1-247; Pars II, 1-256. [see part 2 

p. 27]. 

[Vol. 42, 1955] 


Explanation of Plate 


Fig. 1. Ephelidium heardensis Dodge & Rudolph. 

A. Cross-section of thallus. 

B. Cross-section of orthidium. 

C. Conidiophores. 

D. Conidiospores. 

Fig. 2. Acrospora Gwynni Dodge & Rudolph. 

A. Cross-section of apothecium. 

B. Cross-section of thallus showing spermogonium 

C. Ascus. 

D. Spermatiophores. 

Fig. 3. Collemopsidium pyrenuloides Dodge & Rudolph. 

A. Ascospores. 

B. Algal colonies. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 15 




Ann. Mo. Bot. Gard., Vol. 42, 195 5 

Plate 16 





[Vol. 42, 1955] 



Fig. 1. Heard Island, Laurens Peninsula. 
Fig. 2. Heard Island, Coast of Atlas Cove. 

(Photographs taken by the Australian National Antarctic Expedition.) 



Anaxagorea Allenii R. E. Fr. n. sp. — Arbor circ. 6 m. alta; ramuli novelli 
brevissime rufo-tomentelli. Foliorum petiolus 10-12 mm. longus, crassus; lamina 
papyracea, (exsiccata) flavo-viridis, supra glaberrima laevisque, subtus demum 
glaber et pallidior, oblongo-lanceolata, basi rotundata, apicem versus sensim long- 
eque angustata, 30-35 cm. longa et 8-10 cm. lata; costa supra subplana, subtus 
valde prominens, teres, rugosus; nervi laterales I. utrinque circ. 15, sub angulo circ. 
80° exeuntes, deinde sursum curvati, sicut venulae subtus paulo elevatae. Pedicelli 
florum breves (3-5 mm. longi), crassi, sicut sepala et petala rufo-tomentelli, juxta 
calycem bractea vaginaeformi instructi. Sepala ovato-orbicularia, apice rotundata, 
non recurvata, 7-10 mm. longa. Petala crassa, exteriora oblonga, apicem rotund- 
atum versus haud angustata intus apicem versus carinata, ad 12 mm. longa et 5 


medium intus carinata. Stamina linearia, 3 mm. longa; appendix connectivi tenuis, 
circ. 0, 7 mm. longa, apice rotundato-truncata. (Fructus ignotus). 

Canal Zone: Quebrada Lopez, alt. 30 m., in flower, Feb. 11, 1940, P. H. 
Allen 2143 (Type in Herb. Missouri Bot. Garden) . 

This species is very like the Anaxagorea crassipetala Hemsl., found in Guate- 
mala and Nicaragua, but differs from this especially in its leaves being larger, 
rounded at the base, and with much-elongated, almost thread-like apices; in its 
pedicels being short and thick, in its sepals being longer and of rounded oval, not 
triangular-shape, and its outer petals longer and not tapering toward their apices. 
it is also similar to Anaxagorea dolichocarpa Sprague and Sandwith, from which it 
differs by sepals rounded at the tips and not retroflexed, short inner petals, shorter 

stamina, etc. 

Ramuli novelli pilis brevibus fer- 
tn^^^^^prtsm Y&^U. Foliorum petiolus teres vel supra planus tomen- 





glabrescens, oblongo-oblanceolata, supra medium latissima, basi breviter acuta, 
apice abrupte contracta et 1-1.5 cm. longe cuspidata, 15-20 cm. longa et 4-5 cm. 
lata; costa utrinque elevata; nervi validiores utrinque circ. 15, arcuatim adscen- 
dentes et 1-2 mm. a margine conjunct!, tenues et supra elevati; reticulum 
venularum in foliis adultis haud conspicuum. Inflorescentiae ex axdhs f 
delapsorum evolutae; pedicelli florum crassi, ferrugineo-sencei breves (7 mm. 
longi vel ultra?), paulo supra basin articulati. Sepala inter se libera, rotundato- 
ovata, acutiuscula, circ. 2 mm. longa. Petala rigida, exsiccata nigrescentia, ex- 
teriora rotundata et extra ferrugineo-sericea, circ. 7 mm. in diam., intenora minora, 

medium ferrugineo-tomentosa. Stamina 1.8-2 mm. longa, 


issued June 23, 1955. 


[Vol. 42 


apice truncata. Ovaria 1, 5 mm. longa, prismatica et breviter strigosa, stigmate 
globoso-ovoideo coronata; ovula 3—4, lateralia. (Fructus ignotus). 

Panama: Cerro Campana, trail from Campana to Chica, alt. 600—800 m., with 
flowers Aug. 10, 1941, Allen 2647 (Type in Herb. Missouri Bot. Garden). 

The material is unfortunately scanty, with no fruits at all. The number and 
lateral positions of the ovules, however, prove it to be well differentiated from the 
Central American species U. Pittieri Saff. and Schippii R.E.Fr. Flowers of the 
third and last Central American species, costaricensis R.E.Fr. (Acta Horti Bergiani 
13:105. 1941) are unknown so far, but its fruit structure with one single 
oblique to horizontal seed favours the view that, structurally, the flower agrees 
with that of Pittieri-Schippii, and accordingly differs from panamensis. Unonopsis 
panamensis also differs from costaricensis in having much larger and differently 
shaped leaves. In my survey of the species of this genus (Acta Horti Bergiani 
12:237) panamensis should be ranged among the guatteriodes y obovata, and 
allied species, but cannot be joined to any of these. 

Guatteria inuncta var. caudata R. E. Fr. nov. var. — Dif f ert a typo f oliis 
pro rata angustioribus, basi longe cuneato-angustatis, apice acumine 2-3.5 cm. 
longo caudiformi acutissimo ornatis, pedicellis florum tenuioribus. 

Bocas del Toro, vicinity of Chiriqui Lagoon, Old Bank Island, von Wedel 
2108 (Type in Herb. Missouri Bot. Garden) . 



The recent essay of Jenkins 1 serving as an introduction to Auguste de St. 
Hilaire's Esquisse de mes voyages au Bresil et Paraguay focused considerable atten- 
tion on this French naturalist who collected plants and animals in southeastern 
Brazil during 1816-1822. Few of the 30,000 plant specimens collected by St. 
Hilaire 2 have found their way into other botanical institutions from the herbaria 
of the Museum National d'Histoire Naturelle in Paris or the Universite de Mont- 
pellier 3 . This has contributed to the failure of many modern monographers of 
Brazilian plants to cite St. Hilaire material. 

While studying at the Museum National d'Histoire Naturelle in Paris I came 
upon the ten handwritten books constituting the catalogues of plants collected by 
St. Hilaire 4 . It is my hope to demonstrate the importance of these catalogues and 
also to rectify some errors made in the handling of the St. Hilaire collections. 

I wish to express my thanks to Prof. H. Humbert, Directeur Laboratoire de 
Phanerogamie, Museum National d'Histoire Naturelle, Paris, France, for permission 
to examine and to have microfilmed the catalogues of Auguste de St. Hilaire, as 
well as to use the facilities of the Herbarium. I owe especial gratitude to 
Monsieur J. Leandri, Sous-Directeur of the same institution, for his substantial 
assistance. To Mr. George Swett, Examiner of Questioned Documents, Clayton, 
Mo., I extend my heartfelt appreciation for his comparison of handwritings found 
in and related to the catalogues. 


St. Hilaire set foot on Brazilian soil at Rio de Janeiro on June 1,1816. He was 
destined to be the first French naturalist to penetrate into the interior of south- 
eastern Brazil. Rio de Janeiro, the infant capital of Portuguese Brazil, was from 
the time of Vellozo in 1780 the hub of botanical activity in South America. 
Three years before the arrival of St. Hilaire, Georg von Langsdorff, a German, 
found the capital an ideal center for his botanical work. Two English collectors 
who botanized in Rio de Janeiro and in Sao Paulo, Brazil, James Bowie and Alan 


1 Jenkins, A. E., in Chron. Bot. 10 1 :5-4L 1946. 

2 Rapport sur le voyage de M. Auguste de Saint -Hilaire dans le Bresil et les Missions du Paraguay, 

lu a 1'Academie Royale des Sciences, de Jussieu Rapporteur. 8 pp. Paris, 1823. This rare reprint may 
be found in the library of the Missouri Botanical Garden, St. Louis, Missouri, as well as the Oliveira 

Lima Library, Catholic University, Washington, D. C. 

3 In America, for example, there are less than 100 sheets of St. Hilaire material deposited in the 
Smithsonian Institution (U.S. National Herbarium), Washington, D. C, and less than 40 sheets 

in the Missouri Botanical Garden, St. Louis, Mo. 

4 1 wish to express my gratitude to the National Science Foundation, Washington, D. C, for the 
grant of money to defray in part the expenses of travelling to the Vlllth Botanical Congress held 
in Paris in the summer of 1954; at the Museum d'Histoire Naturelle I had the opportunity to ex- 
amine the catalogues first hand. I prefer to speak of the books as catalogues, despite the fact that it 
may be more accurate to designate the ten books collectively as a catalogue. 

* Missouri Botanical Garden and St. Louis University. 



[Vol. 42 

Cunningham 5 , in their unpublished (Brazilian) diary speak of visiting "Mr. Langs- 
dorfPs garden" on January 10, 1815, Four fellow-countrymen of Langsdorff 
had already penetrated into the interior of southeastern Brazil: Sellow, von 
Chamisso, Eschscholtz, and Maximilian, Prinz zu Wied-Neuwied. The most illus- 
trious German systematist of this era, von Martius, was soon to arrive (July, 
1817). Urban's account of St. Hilaire's five "voyages" informs us that St. 
Hilaire met two of these botanists, Langsdorff and Sellow 6 . 

One of the most concise and accurate summaries of the travels and activities 



travels presented to the French Academy by Baron Cuvier and five other members. 
This represents the acceptance by the Academy of St. Hilaire's Apergu d y un voyage 
dans Vinterieur du Bresil of the previous year 7 . I have taken the liberty of trans- 
lating a section of this "summary" from the French. Wherever possible, the 
spelling of geographical names has been changed to that of modern usage. 
After speaking of St. Hilaire's arrival in Brazil we read: 

In the following December he proceeded northwards, in the province of Minas [Gerais], 
a territory as vast as France, where he sojourned for 15 months in order to explore its various 
parts. At first he headed in the direction of Villa Rica, penetrating into old and very 
extensive forests, covering in turn open country with a different flora, with different birds 
and insects. There he discovered a "quinquina" unlike the Peruvian kind, but used in the 
same way, and serving as an indicator of iron. Likewise he observed a grass sticky to the 
touch, growing luxuriantly in habitats formerly covered with virgin forests. After an illness 
of a month's duration in Villa do Principe, he travelled through other forests, situated to the 
east of this town and studied the survivors of several ancient Indian tribes. He noted in the 
"district" of Minas Novas a change in the flora, with woods composed of only small shrubs. 
Further on, other trees of greater stature appeared, called Catingas, which are leafless during 
the dry season. He spent fifteen days on the banks of the Jequitinhonha in the midst of the 
Botocudos, warlike Indians who live in the forests, naked and without definite dwelling 
places. From there he turned toward the mighty San-Francisco river, followed along its 
banks and entered the District of the Diamonds [Diamantina] and visited the areas where 
he had previously noted their methods of extracting gold from the mountains and rivers. 
Returning to Villa Rica by a long detour, he again headed for Rio (de) Janeiro arriving in 
March of 1818, whence he sent to the Museum d'Histoire Naturelle a shipment of 200 birds, 
some quadrupeds, 800 species of insects, and two hundred packets of seeds, as well as two 
botanical papers which have been printed in the publication of the Museum. 

A second voyage of several months duration brought him to the maritime districts, to the 
north of Rio de Janeiro, as far as the Rio Doce, and yielded just about the same materials 
as those collected in the woods adjacent to the capital with the exception of those markedly 
different ones found on the sandy expanses. He visited Cabo Frio, the town of San Salvador 
de Campos, the entire province of Espirito Santo, the unhealthy banks of the Rio Doce, and 
having embarked at Villa da Victoria, he returned by sea to Rio de Janeiro, whence he sent 
a new shipment of animals and of seeds to the Museum. 

His third voyage, the longest and most important, begun in January of 1819, was at first 
toward the northwest in the eastern section of the province of Minas [Gerais] where he saw 
a beautiful waterfall at the source of the Sao Francisco, and the sulphur springs of Araxa 
whose waters cattle drink unhesitatingly. To arrive at Goiaz he followed a barren plateau 

5 Bowie and Cunningham were to depart from Rio de Janeiro in October of 1816, the former to 
collect at the Cape of Good Hope, the latter in Australia. Their unpublished Brazilian journal is at 
the Royal Botanic Gardens, Kew, England. 

6 Urban, I., in Martius* Flora Brasiliensis l 1 :93-98. 1906. 

7 St. Hilaire, Auguste de, in M£m. Mus. d'Hist. Nat. 9:307-380. 1822. 




whose opposite sides form the sources of the Sao Francisco river and the Dos Tocantins both 
entering the sea from different directions. The vegetation struck him as being little different 
from that seen on the banks of the Sao Francisco. The pastures were at times entirely 
barren, at times covered with small stunted trees. M. de St-Hilaire after having visited Villa 
Boa, capital of the province of Goiaz, spent a few days among the Coyapos Indians, then 
went to the border of Matto Grosso; from here he took a southeastern route to Sao-Paul. He 
arrived there in December 1819, after a long and arduous journey marked by great heat 
and dryness, and here he deposited his various collections, packed with care, to permit him 
to continue southwards. 

At first he skirted the western side of the vast chain of mountains which separate the 
interior of the country from the seashore, and saw in succession the following villages: 
Sorocaba, Itu, Porto Feliz and their environs; he noted the boundaries of the most important 
colonial industries. Then he entered Campos Gerais, such delightful country because of its 
diversified countryside. He made a goodly collection of plants which indicated a more 
temperate climate. There he observed the erroneously named her be du Paraguay; this he 
correctly identified as a tree of the genus Ilex. He also identified the local quinquina which 
is in reality a Solanum. Crossing over a chain of mountains at a difficult spot called Serra 
de Paranagoa, he found himself at the sea coast at a much lower altitude and hotter than 
the interior plateau whose crops approximate more those of the tropics even as far south as 
27° latitude. He was able to visit the islands of St. Francis and of St. Catherine and their 
whaling-posts. Here the ground beyond is a sandy and dry stretch enlivened only by the 
presence of a prodigious number of aquatic birds. 

The province of Rio Grande, which he soon entered, is very fertile, with a more vigorous 
population. He observed that the culture of sugar-cane stops at about 30° while cotton- 
growing extends to 31°, palms to 34°, just at noted in New-Holland [Australia]. He spent 
the winter, already severe in these areas, at Porto Alegre and at Rio Grande, departing in 
October 1820, that is in the springtime of this country. The plants which flourish at this 
time belong to the same genera as those constituting the spring flora in Europe, such as the 
carex [sedges], anemones, centunculus [chatfweeds], arenaria, etc. He recognized, too, the 
influence of climate on vegetation: thus, in the coldest season, the trees retained their foliage, 
up to a degree north of Porto Alegre; a third lose their leaves at Rio Grande, and at two 
degrees further south scarcely a tenth have retained their leaves. 

M. de St. Hilaire then entered the Spanish possessions, passed through several towns and 
beautiful estates along the banks of the Rio de la Plata, arriving at Montevideo where the 
vegetation appeared to him to be so closely related to that of Europe that he scarcely found 
fifteen plants which could not have been assigned to European families. He also noted around 
certain towns that several of our own plants were well acclimated; he called attention to this 
especially near Montevideo, where vast waste-lands were covered with our cardan [thistle] 
growing wild and used solely as fuel. Proceeding then in the direction of the Rio Negro, a 
river rushing headlong into Uruguay, he found beyond a country considerably less populated 
and with customs agreeing with the descriptions of Azzara. His journey became more dif- 
ficult between Belem and the province of the Missions, and he spent thirteen days in an 
uninhabited desert, populated only by autruches [rhcas], cerfs [deer], and jaguars, against 
which he had to defend his own life, and especially that of his horses; unfortunately he 
could not save all of the horses. It was in this district that curious accidents befell him 
and his two companions after having eaten several spoonfuls of a honey made by a species 
of wasp. He visited the seven small villages which are all that remain of thirty which 
constituted in time past the ancient missions of Paraguay. Observing these carefully, he 
could naught but groan at finding only ruins and the sight of such misery in a fertile land 
whose former splendor our writers have not exaggerated. Then re-entering Brazil proper and 
traveling in the rainy season, he frequently ran the risk of losing his notes and collections; 
he returned at the end of a year by a detour to Porto Alegre, embarked on Lake Patos for 
Rio Grande; returned by sea to Rio Janeiro, where he arrived toward the end or 1821, 
after an absence of nearly three years. 

St. Hilaire has elaborated on his travels in his own Esquisse (loc. cit.) and his 
Voyages. These are in need of careful study by contemporary scholars. 

The recent remarks of the Brazilian P. R. Reitz (Anais Botanicos do Herb. 
Barbosa Rodrigues 1:86. 1949) on St. Hilaire reflect the warm affection of the 
Brazilians for the French naturalist. I have taken the liberty of translating freely 
the same from the Portuguese: 

[Vol. 42 


He had a kind heart, adapting himself to the environment in which he lived and like- 
wise to the uneducated people with whom he often traveled, treating with the manners of 
a gentleman those who helped him, censuring amiably those who treated him badly. 

He described objectively the customs of our lands. St. Hilaire stands in marked contrast 
to those pseudo-investigators who, as tourists, describe our country as a grand carnival from 
the seats of gambling halls or even from the summit of Corcovado 8 where they sit to write 
their observations, their memoirs of the people, of the Brazilian way of life and customs. 
Only a St. Hilaire, a Martius or a Bonpland who came to Brazil with great sacrifice, who 
traveled through inhospitable areas, could sense the pulse of the Brazilian heart not only in 
the cities but in the back country and small villages, could see in their habitats great 
diversity of plants, animals, and minerals. Only these or others of their calibre could 
describe, in the role of investigators, our natural riches and our people. 

There is an excellent map of the itinerary of St. Hilaire's five "voyages", pre- 



Lntos for Chronica Botanica (loc. cit. p. 12), from St. Hilaire's 
back of his Voyage a Rio Grande do Sul. A large map found in the 
first volume of Martins' Flora Brasiliensis labelled "Itinera Botanicum" with inked- 
in itineraries of the majority of the principal botanists who collected in Brazil from 
1780 to 18 52 is unfortunately inaccurate, at least with respect to St. Hilaire. For 
example, the delineator indicates that St. Hilaire collected as far north on the 
eastern coast of Brazil as Villa de Porto Seguro (Bahia) ; this is almost 225 miles 
north of the Rio Doce, the most northeastern limits of St. Hilaire's collections. 


The catalogues are made up of separate books in St. Hilaire's own hand- 
writing. Approximately 8900 collections are entered in the 1575 pages 9 . My count 
from the catalogues stands in marked contrast to Urban's statement (Martius 9 
Flora Brasiliensis l 1 :92-98. 1906) of the number deposited in three herbaria, that: 
"Collectio princeps (cr. 7600 numeri) in herbario musei historiae naturalis 
Parisiensis. Dupla in Montpellier, parca in museo Berolinensi". Urban's estimate 
may have been the result of adding up the collection numbers in the catalogues 
without considering that often single collection numbers were subdivided one to 
several times, or from St. Hilaire's own statement (Esquisse, p. 61) that: "Le 
nombre des plantes que j'ai recueillies s'eleve a environ sept mille ....". In the 
catalogues there are approximately 835 collection numbers which are subdivided 
a la "bis", rt ter", "4", etc. In one instance (in series C 2 ) the collection num- 
ber 2791 is subdivided thirteen times and applied to eleven different families of 
angiosperms. The failure on the part of monographers or copyists to add "bis", 
"ter", etc. to a specific collection number has caused considerable confusion in the 
citing of specimens. 

Following each collection number is (usually) the specific plant family (written 
in French) to which the collection belongs. Occasionally only the generic name 
is given. For approximately 550 collection numbers St. Hilaire fails to supply any 
taxa although usually one finds an "analysis" or a few words. In more than one 

8 A high peak overlooking Rio de Janeiro easily reached in our times by street-car. 

9 In a letter recently received, M. Leandri writes: "Our catalogue of entries bears actually 7692 
gatherings [collections] from St. Hilaire; maybe the others have been lost or kept by A. de Jussieu 
and Cambessedes (and others)." 




third of the entries a detailed description ("analysis") of the plant is given in 
French. In some families, for example in the Orchidaceae, these descriptions may 
take up three full pages. St. Hilaire paid particular attention to listing the com- 
mon names of the plants as well as their economic uses. The place of collection is 
usually found at the end of the "analysis". 10 The "analyses" are models of pre- 
cision and critical observation. St. Hilaire himself states (Esquisse, p. 61) 

je les ai toutes analysees sur les lieux-memes, et me suis principalement 
attache a la dissection des parties dont la connoissance repand le plus de lumieres 

sur les rapports naturels". 

Dreuzy, a great-nephew of St. Hilaire, in an appendix to the Voyage a Rio 
Grande do Sul, refers to the difficulties encountered by Saint Hilaire in preparing 
his "journal botanique", the name by which Dreuzy apparently designates St. 
Hilaire's diary and his catalogues collectively: 

Nous donnons ci-dessous quelques extraits de ce volumineux journal pour faire comprende 
au lecteur quel travail s'imposait l'auteur au soir de chacune de ses fatigantes journees de 
voyage, assis sur ses malles sous un rancho, eclaire par la lueur tremblante d'une chandelle 
fumeuse, devore par les moustiques qu* attirait la lumiere. 

Dreuzy then supplies two samples of St. Hilaire's "analyses" from catalogues 
C 2 , Section 2, one collection number 2682, and the other 278 f l \ To my knowl- 
edge these are the only "analyses" which have been published in toto. In fact, I 
have found no reference of any length in the literature concerning St. Hilaire's 

mav use Dreuzv's sample 2682 as typical of an "analysis" of 

catalogues. We 
St. Hilaire: 11 

Estancia de Santo-Rei, 1" mars, n° 2682. Graminee. Sur chaque dent de 1'epi sont 
2 fleurs, Tune sessile, l'autre pediculee. Epillets sessiles, 1 fl. herm., lanceoles, sous-falciformes, 
glumes 2-valves, a valves ext. herisee, lanceolee, tantot simplement aigue, tantot courtement 

1 ou 2 arretees a valv. int. un peu plus courte que l'autre lanceolee-aigue, pubescence. Glum- 
elles 2-valv., a valv. membraneuses sous-violettes, sous-pubescentes presque egales, obtusiuscules, 

2 paleoles tronquees; 3 et 2 styles setaces, 2 stig., complets, continus, plumeux, ovaire glabre 
ovoide. Epillets pedicelles males. Glumes 2 valv. a valve ext. pubescente, lanceolee-lineaire- 
aigue, plane, a valve int. lanceolee-aigue, presque egale a l'autre pubescente carenfce, 2 
paleoles tronquees. 2 et. glumelles 2 valv. a valves lanceoUes, hneaires, aigues, presque 
egales, ciliees, membraneuses, un peu violettes. Toute la plapte repand une forte odeur de 
citron. Nom. vulg. capim Ihnao. 

Unfortunately, St. Hilaire made no great effort to list binomials in the cata- 
logues. In the more than 8900 collection numbers there are approximately 160 
binomials, the majority in St. Hilaire's handwriting and a few in Spach's. Most of 

10 The catalogues do not list as many specific localities as does Urban in his elaborate list of the 
itineraries of St. Hilaire in Martins* Flora Brasilivnsis (be. at). Here more thar ,800 local- 
given arranged in proper sequence as well as according . tc ; the • J-- ^^ "^es five 

ities are 

"voyages". It is probable that Urban gleaned this detailed itinerary from St. s own pub- 
lished reports of his trips. 

"Unfortunately, in 2682 Dreuzy terminates this "analysis" of a Graminee with the binomial 
Ecbites guaranitlc* Aug. de S.-H. This binomial, despite the fact that ,t appears ,n the catalogues 
to be written into this "analysis" at the upper right-hand corner obviously refers to co lee .on 
number above, 2681 (Apocynacee). To make matters worse, the second sample analyse 1S 
labelled "Composee 2787**'." From the catalogues the collection number is simply Z787. 

[Vol. 42 


the binomials are addenda to the "analyses"; those written by St. Hilaire belong as a 
rule to families on which he published extensively: e. g., Violaceae (including cer- 
tain genera now assigned to the Ochnaceae) , Droseraceae, etc. For some reason 
"analyses" of certain families, e. g., Malvaceae, Malpighiaceae, etc. are often cross- 
hatched in ink. In the case of the Rutaceae the numbers of the family were 
designated by a scrawled name which looks like "Oranger" (pi. 17, under 2?QI 12 ). 
Figures 1 and 2 show the number of collections made in specific families (with 
the exception of the ferns) according to the catalogues of St. Hilaire. Figure 1 
lists the families represented by more than 50 collection numbers; fig. 2 lists those 
represented by less than 50 and more than 20 collection numbers. The family 
Leguminosae is considered in sensu lato to include the Mimosaceae and Caesal- 
piniaceae. The Ochnaceae includes certain genera which St. Hilaire related to the 
Violaceae, e. g., Sauvagesia, Lavradia. To avoid confusion I have combined the 
Liliaceae with the Amaryllidaceae. 


In preparing a collection of plants for herbarium deposit the most convenient 
sequence of collection numbers is a continuous linear one which is correlated with 
the date of collection and the itinerary of the collector. In examining the more 
than 8900 collections of St. Hilaire we find the following numerical sequences 12 
in the ten books: 

Book 1 — 1- 758 Book 6 — 86- 96 

Book 2- } A f+2 f??n Book 7- ) *~ 60 < not in Su Hilaire,s handwritin S> 

Book 4 
Book 5 




■ 650 




63, 103- 







• 408 

Book 3— 1- 63,103-2085 $1176-1899,2000-2797 

Book 8-j 1-219 

Book 9 — 692- 709, 1858-1869, 2004-2120 
Book 10 — 1- 818 

Despite the patch-work pattern of these sequences of numbers definite cor- 
relations exist between sets of numbers and the itineraries of St. Hilaire's five 
voyages. It is not impossible that St. Hilaire prepared the catalogues following 
his return to Paris in 1822. I have seen certain individual "analyses" attached to 
herbarium sheets in the Paris Herbarium, e. g. to one of several of St. Hilaire's un- 
numbered collections of Lavradia capillaris (Ochnaceae). It may be that the 
"analyses" were written on separate sheets of paper in the field, inserted separately 
with the unmounted specimens and then removed in Paris to be gathered together 
and recopied into a catalogue. The mounting of his 30,000 specimens in Paris may 
have resulted in some confusion in the rc-organization of the "analyses" in the 
catalogues. Occasionally, one finds the date 1822 at the end of an analysis, sug- 
gesting that an addendum was made to the original more-abbreviated "analysis". 
The "analyses", too, arc generally written in a neat (but often difficult to decipher) 

12 In a given run of a number in the catalogues it is not uncommon to find occasional numbers 
missing or out of place. For the sake of brevity I am not specifying these, although the absentees 
have been taken into account in computing the sum-total of collection in the catalogues. 


























100 200 500 400 500 600 700 800 900 1000 1100 

Fig. 1. Approximate numbers of collections 
(except for Ferns) in the botanical catalogues of 
collections in excess of 50 are shown. 

listed under specific 
Aug. de St. Hilaire. 

families of Angiospermae 
Only those families with 





























Fig. 2. Approximate numbers of collections listed under specific families of Angiospermae in 
the botanical catalogues of Aug. de St. Hilaire. Only those families with at least 20 and not more 
than 49 collections are shown. 

[Vol. 42 


hand. Perhaps legibility is too much to expect of a scientist working in the 
evening after an exhaustive day in the field. 

In Book 5 Weddell writes that certain notes for collection numbers 2085-2101 
were not sent to the Museum for insertion. Several points argue against this 
possibility, however. In the first place, it is more logical, if not more safe, to 
insert the "analyses" initially into a field book; again we note, especially in the 
early books, that St. Hilaire often fails to complete an "analysis" before beginning 
the next collection number but occasionally completes the "analysis" on another 
page and consequent to a different collection number. It is difficult to believe that 
he would do this if he were preparing the catalogues in Paris. The fact that 
volume 6 has only 1 1 collection numbers sandwiched in with what is obviously 
a segment of his diary written in the field is another substantial point of evidence. 

St. Hilaire and his contemporaries used these numbers in the catalogues of col- 
lections and for herbarium sheets, and not for purposes of citation in the literature. 


'dionalis, Jacque 

No doubt, they 

did not consider individual collections to be as important as do our botanists of 
today who adhere to the principles of the type concept. 

Obviously, if a collection of St. Hilaire had been cited, e.g.,No. 26, the reader 
would wonder to which sequence to assign it unless the locality were given. Wed- 
dell 13 , no doubt stimulated by the pioneer efforts of certain authors in Martins' 
Flora Brasiliensis in citing collectors and collection numbers, attempted (probably 
in the 1860's) to organize the catalogues in a manner more suitable for citation 
and for labelling of specimens. Unfortunately, Weddell did not publish his "Key" 
to the catalogues 14 . The facts that Spach, who was curator of the herbarium at 
the Museum d'Histoire Naturelle, employed Weddell's system on printed labels 
and that certain monographers have used the Weddellian system give some status 

to the key. We shall discuss the popularity of Weddell's key in following para- 

13 H. A. Weddell (1819-1877), an "aide-naturaliste" to Adrien de Jussieu, collected extensively 
in Brazil; in addition he was a distinguished taxonomist at the herbarium of the Museum d'Histoire 
Naturelle, Paris. 

14 In recent years some members of the staff of the Museum d'Histoire Naturelle, Paris, prepared 
a typewritten copy of Weddell's key. Weddell's handwritten key may be found scattered in the 
front of the catalogues of St. Hilaire (see pi. 18). An accurate handwritten summary (not figured 
here) of the sequences of numbers (accompanied by broad geographical data) of St. Hilaire's 
botanical collections occurs in the initial two pages of Book 8 of the catalogues. Mr. Swett, Ex- 
aminer of Questioned Documents, Clayton, Missouri, expresses the probability that this was prepared 
by a person (or persons) other than St. Hilaire, Weddell, or the author or authors of the page 
shown in pi. 19. It was obviously not prepared by Spach. I have elected to credit Weddell with 
the establishment of the key to St. Hilaire's collections. That Weddell inserted notes in numerical 
sequences and geography in many of the books of St. Hilaire seems certain. M. Leandri of the 
Laboratoire de Phanerogamic Museum D'Histoire Naturelle, Paris, writes me that the typewritten 
key was^copied from notes in the books [St. Hilaire's catalogues] partly of the handwriting of 
Weddell". While Weddell may have based his insertions on the well-organized summary found in 
book 8, it seems safer to suppose that the latter was prepared from his notes. 




Weddell divided the book 

series: A (books 1, 2, 3, and 4) ; B (books 5 and 6) ; C (books 7, 8, and 9) ; and 
D (book 10). Series A is divided into two volumes, B, into two volumes, C, into 
three volumes, and D consists of one volume. The following revised outline is 


at Paris: 



A*j comprising numbers I— 758. 

Plants collected during the year 1816, some in the vicinity of Rio de Janeiro (1—469 and 651- 
758), others in the vicinity of Uba, on the Rio Paranhyba about 25 leagues from Rio de Janeiro, 
and in the same province (470—650). 

A 2 , in two sections. 

Section 1, comprising numbers 464-650, which are duplications of the same numbers in 
Volume A 1 . 

Section 2, comprising numbers 1 B -130 B (or 132 B ). Plants from the environs of Rio de Janeiro 
collected in 1818. 


B 1 , comprising numbers I—2085 (December 1816 to March l8l8j. — 

Plants collected in the province of Minas Gerais, with the exception of 1-25, which are from 
the province of Rio de Janeiro. 

B 2 , in two sections. 

Section 1, comprising the numbers 2101 b,s to 2493; these follow in order the series oi plants 
of Minas Gerais which constitutes the three previous volumes. 

Section 2, comprising the numbers 1-408. Plants collected during a "voyage" on the seacoast 
of Brazil, from Rio de Janeiro to the Rio Doce in the provinces of Rio de Janeiro and Espirito 
Santo, from August to November, 1818. 


C 1 , comprising numbers I-II75. 

Plants collected in the following districts: 

1- 10 — Province of Rio de Janeiro. 1819. 
11- 625 — Province of Minas Gerais. 1819. 
926- 985 — Province of Minas Gerais. 1819. 
626- 920— Province of Goiaz. Sept. 1819 to May, 1820. 
986-1175 — Province of Sao Paulo. Sept. 1819 to May, 1820. 

C 2 , divided into two sections. — 

Section 1, comprising numbers 1176-2797. Plants collected in the following regions; 

1176-1704 — Province of Sao Paulo. May, 1820. 
1705-1806— Province of Sta. Catarina. May to June, 1820. 
1807-1874— Province of Rio Grande do Sul. 1820. 
2569-2797— Province of Rio Grande do Sul. 1820. 

2000-2260 I - From San Pedro de Rio Grande to Montevideo. 1820. 
2261-2568 — From Montevideo to the Brazilian frontier on the banks of the 

Uruguay. 1820. 

Section 2, comprising the numbers 1-2 19c. 

Plants collected in the environs of Rio de Janeiro, in July, 1821. 

Comprising three small series of numbers (692-709, 1858-1870, and 2004-2020) 
tercalated in the large series of numbers contained in volumes C and C . 

15 Weddell referred to these as volumes. 

[Vol. 42 




D, comprising the numbers 1—818. — 

Plants collected in 1821 and 1822 during the last voyage of St. Hilaire, in the following places: 

1_ 30 — Province of Rio de Janeiro. 
31— 587 — Province of Minas Gerais. 
588- 818 — Province of Sao Paulo. 

Certain elements of confusion exist in the interpretation of these numbers. In 
Book 2 the numbers 464-650 are simply duplicates ("double emploi") of the 
same numbers in the 1—758 sequence of Book 1. The numbers 1—112 in Book 4 
are duplicates of those found in Book 3 except for 64-102 which are missing in 
Book 3. The numbers 86-96 in Book 6 (if one should refer to this as Book 6 in- 
asmuch as it is apparently one of St. Hilaire's diaries) belong to the sequence of 
numbers in Book 5 from which they have been omitted. In Book 7 the initial 
sequence of numbers 1—60 are not in St. Hilaire *s handwriting 16 . In Book 8 num- 
bers 1900-1999 are missing, apparently due to carelessness. 

Weddell (?) used paper labels for all the specimens of the collections except 
numbers 1-758 in catalogue A 1 (Rio de Janeiro), 1-25 in B 1 (Rio de Janeiro), 
and 25-2085 in B 1 (Minas Gerais), for which parchment labels were used. All 
paper labels, except for numbers 2101 b! " to 2469 of B 2 section 1 (Minas Gerais), 
and 2470 to 2493 B 2 section 1 (Rio de Janeiro) bear, or should bear, symbols. This 
is summarized as follows: 






Paper b 

1- 240 

B 2 Sect. 2 

Rio de Janeiro 

241- 408 

B 2 Sect. 2 

Espiritu Santo 

Paper B 

1- 130 

A 2 Sect. 2 

Rio de Janeiro 

Paper C 

1- 219 

C 2 Sect. 2 

Rio de Janeiro 

Paper D 

1- 30 


Rio de Janeiro 

31- 587 


Minas Gerais 

588- 818 


Sao Paulo 

Paper O 

1- 10 

C 1 

Rio de Janeiro 

11- 625 

C 1 

Minas Gerais 

626- 691 

c 1 


692- 709 

c 3 


710- 920 

c 1 


926- 985 

c 1 

Minas Gerais 


c 1 

Sao Paulo 


C 2 Sect. 2 

Sao Paulo 


C 2 

St a. Ca tar in a 


C 2 

Sao Paulo 


C 2 

Sta. Catarina 


c 2 

Rio Grande do Sul 


c 3 

Rio Grande do Sul 


c 2 

Rio Grande do Sul 


c 2 

Banda Oriental del Uruguay 


c 3 

J— fc 

Banda Oriental 


c 2 

Banda Oriental 


c 2 

1 Rio Grande do Sul 

16 These 60 numbers have neither families nor "analyses" appended, although the localities of the 
collections are given. Some of the localities (e. g., Barbacena, Rio Ouro Branco) suggest the begin- 
ning of St. Hilaire's second voyage to Minas Gerais in 1816. On this voyage he was accompanied 
by I. Gomes and G. H. de Langsdorff. Perhaps these are collections of one or both of these men. 




Although only a few botanists such as Wcddell, Spach, Dreuzy, etc., are 
known to have consulted the catalogues of St. Hilaire, the various check-marks 
and binomials inserted suggest that they may have been studied by others. 

Prior to an arbitrary date of 1860 monographers in citing St. Hilaire collec- 
tions, in imitation of St. Hilaire, Cambessedes, and de Jussieu failed (usually) to 
supply collection numbers but gave clear-cut geographical data. Later mono- 
graphers often used St. Hilaire collection numbers, and, almost without exception, 
they furnished only a fraction of collections as listed in the catalogues. Mueller, 
for example, in his classic work on the Apocynaceae (Mart. Fl. Bras. 6 1 : 1-195. 
1860) gave collection numbers for only 23 of the 56 collections of St. Hilaire 
which he cites. These 56 collections represent about one-third of this family as 
listed in St. Hilaire's catalogues, all of which have specific collection numbers. In 
general, in Martins' Flora Brasiliensis it is the exception to find collection numbers 
when St. Hilaire material is cited; collecton numbers are missing, for example, in 
Radlkofer's monograph of the Sapindaceae (1895), Bennett's work on the Poly- 
galaceae (1874), Koehne's work on the Lythraceae (1887), etc. (see Table II). 

No doubt much of the confusion and inconsistency encountered in the citation 
of St. Hilaire's collections, especially by the 19th century monographers, arise from 
several circumstances: (a) localization of the bulk of the St. Hilaire collections 
in the Paris herbarium; (b) the fact that the duplicate collections often have 
labels bearing only a binomial and the geographical data (often undecipherable) 17 ; 

(c) the lack of interest in collection numbers by the contemporaries of St. Hilaire; 

(d) awkward and overlapping sequences of numbers prior to Weddell's groupings; 
(f ) the subsequent failure to apply the Weddellian system to many of St. Hilaire's 


Table II, based on an examination of twenty monographs (1832 to the present) 

which include St. Hilaire's collections, reveals many important points 18 . In the 
Anacardiaceae (D.C. Monograph. Phanerogam. 4. 1883) Engler cites eleven col- 
lection numbers of St. Hilaire, none of which has catalogue numbers; in attempt- 
ing to check them against the catalogues it appears from the blank spaces in 
the latter that St. Hilaire may not have been certain of the characters of the 
Anacardiaceae and thus never entered the family name. In Cogniaux's Cucur- 
bitaceae (D.C. Monograph. Phanerogam. 3. 1881) the number of St. Hilaire's 
collections cited exceeds the total number estimated to be entered in the catalogues 
under this family. This disparity is a reflection of the number of blanks left by 

17 Of the 59 sheets of St. Hilaire reported by Dr. Lyman B. Smith, of the Smithsonian Institute 
(Washington, D. C), to be in that herbarium only 34 bear collection numbers; of these only two 
have the Weddellian series appended to the number. I have been unable to locate ten of the num- 
bered specimens with certainty in the catalogues. . 

18 When the collection numbers are not listed in a monograph the author may be including more 
than one collection (of St. Hilaire) for the species in question without making the same clear. 
This makes an accurate statistical analysis of the number of collection cited impossible. 



[Vol. 42 

Specific Monograph 


A. de Jussieu 












Rubiaceae Tribes VII-XIX 


Sapindaceae (Serjania) 





Ochnaceae (Ouratees- 

Van Tieghem 



Vochysiaceae (Salvertia & 

Vocbysia ) 

Labiatae (Hyptis) 





















































| .2 













































s c 

•i e 




o — 














— -H 













































■ 1 * 











• M 
































St. Hilaire after collection numbers in the catalogues, some of which are un- 
doubtedly of the Cucurbitaceae. 

In Naudin's monograph of the Melastomaceae 19 , we find very few of St. 
Hilaire's collections cited despite the fact that Naudin must have been well 
acquainted with them 20 . "What is even more remarkable is the complete absence 
of St. Hilaire material in the "second half" of the monograph (i. e., vol. 16, after 
p. 87. 1851), even though such common Brazilian genera as Clidemia, Miconia, 
etc. are included 21 . In the several sample monographs from Martins' Flora Brasil- 
iensis we note that few collections of St. Hilaire are cited as compared with the 
total number of collections of the specific families estimated to be in the catalogues. 

Killip, in his recent treatment of the Passifloraceae (Field Mus. Nat. Hist. Bot. 
19:1-613. 1938), cites six numbers which do not check with the corresponding 
ones in the catalogues. Of these, three (717, 751, and 970) were probably 
erroneously copied from labels bearing the numbers, 719, 753, and 976 respectively, 
all of which are entered in the catalogues under Passifloraceae; a fig. 9 may be 

easily mistaken for a 7, a 1 for a 3, etc. 

Epling, in his monograph of Hyptis (Rev. Mus. de La Plata, n. s. Secc. Bot. 
30:153-497. 1949) cites more St. Hilaire material than any contemporary mono- 
grapher 22 . Unfortunately, only two of the hundred collections cited bear collec- 
tion numbers with Weddell's series added. This suggests, of course, that the 
Weddellian system has not been applied in any great degree to the labelling of 
specimens. Despite the quantity of St. Hilaire material cited by Epling, approxi- 
mately 6 5 collection numbers of Hyptis were not included in his work. In checking 
this genus in the Paris Herbarium 13 of the 92 collection numbers (St. Hilaire) 
cited by Epling, were not located; but eight of these thirteen can be located 
readily when "bis", "ter", etc. are appended. (Some copyist of course may have 
failed to insert "bis", "ter", etc. on the labels.) Of the remaining numbers, prob- 
ably 578 was misread for 518, and 1155 for 1133. 

In those monographs using the Weddellian system the collection numbers 
are usually matched with ease in the proper volume in the catalogue. 

"Ann. Sci. Nat. Bot. Ser. Ill, 12:196-284. 1849; 13:25-39 126-159 273-303 .347-362. ,850; 
14:53-76, 118-165. 1850; 15:43-79, 276-345. 1851; 16:83-246. 1851; 17:305-382. 1852; 
i g • g 5 i 54 258—294. 1852. 

"St. Hilaire and' Naudin were co-authors of a paper entitled "Revue de la Flore du Brttl 
Meridionals which appeared in three sections (Ann. So. Nat. Bot. Ser II, J'- 2 *- 5 *. 20 »""S. 
1842). In addition they were co-authors of a new genus, Augustine*, of the Melastomaceae (Ann. 
Sci. Nat. Bot. Ser. Ill, 2:145. 1844). 

21 1 hope to discuss in another paper the stormy history of the Melastomaceae collected by St. 

Hilaire. , 

"Hyptis is one of the few cases where St. Hilaire wrote a genus name with fair consistency into 
his catalogues in his "analyses" instead of the family name "". However, ,n checking 
Epling's monograph against the catalogues, it is obvious that St. d.d on occas.on use the 
taxon of family in place of Hyptis. 


[Vol. 42 



In the herbarium of the Museum d'Histoire Naturelle, Paris, there exists an 
unpublished list of about 2000 entries of St. Hilaire's collection numbers; these 
are arranged according to genera (occasionally binomials are listed) and cover 
eighteen families or tribes. The author of this list is unknown 23 . Fortunately, 
the collection numbers of St. Hilaire are entered according to the Weddellian 
system. The presence of geographical data increases its value considerably (cf. 
pi. 19). 

In the list the entries from numbers 1 to 364 are of the Graminee only, divided 
according to 1 3 tribes. Among the Dicotyledoneae we find the following families 
and tribes listed in the order given (entry numbers 1-1589): 















85 bl8 -95) 









103 bU -103 5 ) 














150-182 bU ) 



308-322 bls ) 


329-3 37) 

338-390 bU ) 



It is obvious that such a list may be of great value to specialists of the above 
groups working on material from southeastern Brazil As an example of its use- 
fulness, I was able to ascertain in the model analysis (cf. p. 157) that Elionurus is 
the genus of the grass family described by St. Hilaire. 



the collection numbers of St. Hilaire's material with the specific catalogue series 
and section, probably due to the absence of these numbers from the majority of 
the labels of the specimens. The 

The Weddellian system answers the problem of the 
possible overlapping of collection numbers, and monographers should be encouraged 
to adopt it. Being aware of the existence of the catalogues of St. Hilaire, they are 
now able to check collection numbers against St. Hilaire's "analyses" 24 . The cata- 
logues themselves represent an original source of reference for the 8900 collections 
of Aug. St. Hilaire in southeastern Brazil. 

The discovery of a list (author unknown) of some of the genera collected by 

system, should be of great assistance to monographers. 


There are 17 collection numbers of St. Hilaire and accompanying data at the end of the list 
which are obviously in Spach's handwriting. This permits one to give the date of the list as some 
time prior to 1879, the year of Spach's death. Mr. George Swett, from a comparison of one sheet 
of the above list (pi. 19, a photo from microfilm) with photographs of the writings of St. Hilaire 
and Weddell (cf. pis. 17 and 18) expressed the opinion neither prepared the list; he qualifies this, 
however, by stating that "no opinion is held as to whether the writers (St. Hilaire and Weddell) 
executed the word Andropogoneae in the above sample sheet." 

The author, through the Missouri Botanical Garden, will be happy to supply sets of collection 
numbers of St. Hilaire for specific families to specialists. 




Through a letter received during the proofing of this paper I learned that the 
Bibliotheque Centrale (Mus. Nat. Hist. Naturelle) has no manuscripts of Aug. 
de St. Hilaire except several fragments of the Flore du Brisil Meridionale and four 
of his letters. Apparently there are no St. Hilaire manuscripts in the Library of 
the Division de Phanerogamie of the Museum except the catalogues herein described, 
and a single letter. 

In addition, I learned that the editor of St. Hilaire's Voyage a Rio Grande do 
Stil is Charles Roland de Dreuzy (1837-1899), a native of Nivernais. He was 
the grand-nephew of St. Hilaire, heir to his estate at Turpiniere, and grandson of 
a sister of St. Hilaire. At his own expense he published the Voyage a Rio Grande 
do Sul, in accordance with the last wishes of his "parent". The chateau at Tur- 
piniere is located near Ferte Saint-Aubin, Sennely (Loiret). The present owner 
is M. Robert de Dreuzy, son of the editor of Voyage a Rio Grande do Sul. 

The grass described in the model "analysis" (cf. p. 157 of text) was determined 
by Miss Agnes Chase as Elionurm candidus var. bisetosus Hack, and Linden. 


[Vol. 42, 19551 


Explanation of Plate 


A portion of one page of Catalogue C 2 of Auguste de St. Hilaire. This is in his hand- 
writing and covers numbers "2791" Hypoxis, 279 1 10 , Labiee, 279 1 11 , Solanee, 2791", 
Oranger", and "Suite a 2791°." The last is a continuation of the "analysis" of Hypoxis 
at the top of the page. 















■HA* 4iJp*(^«P>pMF PP^^^" - ^ 1 

& 3*4** 


Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 18 

















* — 






fit 6 


















: / 






'/t/j m. j-r - 



"** rtV ^ 4|K*v- «Wr i^ 






[Vol. 42, 1955] 


Explanation of Plate 


A portion of the "key" to the botanical catalogues of Auguste de St. Hilaire. This 
is in Weddell's handwriting and is from Series C, catalogue C 1 . The handwriting at the 
top and bottom of the page is that of the librarian. 


[Vol. 42, 1955] 


Explanation of Plate 


A page from a list of approx'mately 2000 collection numbers of Auguste de St. Hilaire. 
Note that the Grasses are entered according to tribes and genera. The fourth collection 
entered under the genus Elionurus (C 2 No. 2682) is that upon which the "model analysis" 
(see p. 157) is based. Thelist is deposited in the Herbarium of the Museum National 
d'Histoire Naturelle, Paris. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 19 

m . $$& 





'jf'fterm* tta/***t // >£? m m$Wt 

ffijom* "ut^iwcj 


*r*/ «£ <« "t 



/&4H4 Jt4*<4 

vn Suj jcm 
















f0«m <**&■/*** 

t>\ if iRf . ••mttou 

Jm V ' mtm/«u** <^/«m> 

it* t/J4 4 

/r ' ft t 


y ■ ■ 



4' »'**4 

<?* *~t>< 

m- ft 4* 





► ft?c 

C* »• f*3t 






Using comparative morphology as criteria, I presented evidence in an earlier 
paper (Hall, 1952) that Juniperus virginiana L. has differentiated into five geo- 
graphic races, and two of these races, called Typica and Ozark, were described in 
detail. Typica is known to the horticulturist as the Eastern form and may be 
characterized as the very tall, narrowly pyramidal tree found from Virginia to 
southern Illinois. Within its present distribution Typica is concordantly variable, 
and at least so homogeneous that it seems to be carrying no introgressant genes 
from other species. The influence of Typica in the juniper populations of the 
Ozark Plateau is not evident even though isolated plantings of this race are to be 
found on the western banks of the Mississippi River from St. Louis southward, 
usually about old farmsteads. The only other race of red cedar which occurs 
occasionally in the St. Louis area is the Northern, which is characterized by high 
frequencies of the spire-like form known as var. crebra> by crooked fruiting 
peduncles, and a preference for well-drained, sandy soils with a slightly acid reac- 
tion. I postulated that this Northern race may be the consequence of a few genes 
from /. horizontalh becoming stabilized in red cedar. The Northern race, with its 
tendency toward suppressed lateral branching, follows the glacial drift down the 
Mississippi River, and where the periglacial loess occurs these junipers are abundant. 
South of the loess deposits, particularly in the vicinity of St. Louis, juniper popula- 
tions closely resembling the Northern race occur sporadically. 

The Ozark race is indeed the most abundant and important in the St. Louis 
area and represents red cedar throughout the Southwest. The Ozark race, which 
is the result of introgression by genes of /. Ashei Buchholz into /. virginiana L., 
includes a highly variable assemblage of junipers distributed from central Texas 
across eastern Oklahoma and the Boston Mountains of Arkansas through the Mis- 
souri Ozarks to the Mississippi and Missouri rivers. Introgression by genes of /. 
Ashei into red cedar has probably occurred repeatedly, at least during the Pleisto- 
cene, and current mixing is evident in restricted areas where extensive hybrid 
swarms exist. 

The purpose of this investigation was to determine the differences between the 
habitats in relation to the differences in the variability within and between popula- 
tions of these Northern and Ozark races of junipers. The habitats studied are close 
together, less than a mile apart, and are surrounded by many hundreds of popula- 
tions of junipers both large and small. It was thought that these habitats may 
select from this large gene pool adaptive genotypes which may vary from popula- 
tion to population proportionately to the habitat differences. 

* Cranbrook Institute of Science, Bloomfield Hills, Michigan. 



[Vol. 42 



In the spring of 1952 I was granted a leave from the Cranbrook Institute of 
Science to be a guest lecturer in Genetics and Evolution at the Henry Shaw School 
of Botany, Washington University, St. Louis, Missouri, and to take over Dr. Edgar 
Anderson's class during his absence. This study was part of a project undertaken 
by Dr. Anderson's class under my direction. 

I am most grateful to Dean Henry N. Andrews for his advice and cooperation 
and to Dr. Anderson for his enthusiasm in regard to field studies as well as his 
helpful criticism. The study was carried out in the native tracts of the Missouri 
Botanical Garden Arboretum, Gray Summit, Missouri, which is approximately 
forty miles southwest of St. Louis on US Highway 66. I am especially grateful 
to Mr. August P. Beilmann, Superintendent of the Arboretum, for his cooperation 
and detailed knowledge of land use in the Ozark region. 


The areas studied are within the Missouri Botanical Garden Arboretum, which 
is situated 40 miles southwest of St. Louis, Missouri, just south of the junction of 
Highway 66 and Highway 50 near Gray Summit, a hilly region principally of 
weathered dolomitic limestones in the northern portion of the Ozark highland. 
The climate is designated by Thorn thwaite as BB^ (humid, mesothermal, with 
precipitation adequate at all seasons). Records in St. Louis for the past hundred 
years show such variations in precipitation that no general trends are evident, so 
that it is difficult to judge whether climatic amelioration or deterioration is in 
process. The Ozark Plateau is a part of the Oak-Hickory Forest Region described 
by E. Lucy Braun (1950) as the Interior Highlands, Southern Division, which lies 
essentially south of the glacial border. Braun wrote: 

"The Southern Division is characterized by the prominence of southern species of oaks 
and hickories, as Quercus stellata, Q. marilandica, Q. Shumardu (and var. Schneckii), and 
Carya Buckleyi (var. arkansana and var. villosa) , and in the bottomlands, Q. nigra, Q. lyrata, 
Carya Pecan, C. myristicaeformis, and C. aquatica . . . . Prairie openings, limestone glades, 
and balds locally interrupt the forest cover." 

Those who know the Ozark landscape consider these local interruptions of 
forest cover as the particular mark of beauty of the uplands, and the "openness" 
was a major feature in selection for settlement in pioneer days. Visitors to the 
Ozarks from foreign countries usually are most impressed by the red cedar glades 
which are island playgrounds in the present-day dense oak woods. In the proper 
season these open areas serve as edge for much wildlife activity, and perhaps the 




The soils indicate a long history of open conditions where prairie vegetation, 
probably maintained by fire and grazing, played a very important role in the 
development of soil horizons. Soils in the area are characteristic of prairie-forest 



transition regions and are much like the nut-structured prairie-forest Grood soils 
of continental sub-humid steppes, being intermediate between chernozems and 
podsols. Regardless of its history of prairie development, the Ozark region has a 
forest climate and probably has had since the end of the xerothermic period. Thus 
the climax plant formation, which is of course subject to continual change, may 
be the oak-hickory forest which has been spreading extensively throughout Missouri 
roughly in the last hundred years. Beilmann and Brenner (1951) and Etter (1953) 
report this trend in their studies of Ozark vegetation. It is not likely that any 
region has ever had a vegetational mantle entirely in equilibrium with climate 
alone, for other factors affecting balance are legion — e. g. nature of geological 
substrate, grazing and browsing, fire, and man. Rather, as Sauer (1950) has 
pointed out, "Plant associations are contemporary expressions of historical events 
and processes, involving changes in environment and biota over a large span of 
geologic time." However, Braun (1950) stated that knowledge of these historical 
events and processes is very scant even though general trends seem to be quite clear. 

The history of land usage in the Gray Summit area is not known in detail, but 
titles to land were granted beginning about 1850, nearly 50 years after the first 
settlers, mostly French, began to locate along the major waterways in what is now 
Franklin County, and over 80 years after St. Louis was founded, also by the French. 
Evidence of timber cutting, heavy grazing and burning variously from the time of 
settlement until the Missouri Botanical Garden acquired the property in 1925 is 
clear and common. Apparently, in the days before the settlement of the region 
most of the timber was found on the headwaters of the creeks and was scattered, 
giving a prairie- or park-like aspect presumably as a result of a somewhat drier 
climate than today and the effects of grazing, fire, etc. The timber, where dense 
enough to be called woodland, was usually situated in the rough lands or dissected 
headwaters of the streams where it was largely safe from fire. 

Today the Ozarks are well covered with close forests, and the once open areas 
show signs of rapid forest encroachment; old fields develop rapidly to forest, but 
the glades show little invasion by woody species even when protected from fire or 
grazing. Steyermark (1940) made a vegetational survey of Missouri which cov- 
ered fifteen years of extensive observation, and even though the survey was strictly 
qualitative it was detailed enough to permit rough comparisons with present-day 
conditions. He stressed the role of rock strata and soil, i. e., edaphic factors, in 
determining the distribution of plant associations. He described six plant associa- 
tions and many more associes based on differences in soil, slope, and water avail- 
ability; in his conclusions he favors the polyclimax concept subscribed to by Domin, 
Gams, Gleason, Tansley, and Du Rietz. He indicated that the glade flora was the 
result of edaphic factors, and in most sites would eventually support a maple- white 
oak association which in the Ozarks seems to be restricted to slopes and bluffs with 
neutral to alkaline soils underlain by limestone. 

The Glade. — The Glade, which is one of several within the boundaries of the 
Missouri Botanical Garden Arboretum, may be characterized as a small open space 


[Vol. 42 

(open glade), approximately 60,000 square feet in extent, covered by grasses and 
junipers surrounded by a closed area (closed glade), or buffer, approximately fifty 
feet in width, chiefly composed of junipers and chinquapin oaks and situated on a 
west-facing slope (upward ENE, 10°20'). The area studied constituted a rectangle 
400 X 150 feet, approximately the total extent of the open glade. In this study 
the term Glade refers to the open glade within the enclosure studied. 

This same glade was described by Erickson, Brenner, and Wraight (1942), who 
demonstrated a direct correlation between the structure and composition of the 
glade flora and the occurrence of a dolomite stratum at or near the surface. The 
glade soil is very shallow (averaging about 4 inches), organically rich and derived 
from a dolomitic rock which under the open glade is a thin, porous dolomite locally 
known as "cotton-rock/' an upper stratum of the Cotter Formation, and lies un- 
conformably on the Jefferson City Formation, both in the Canadian Series of the 
lower Ordovician. Under the closed glade a less porous dolomite with lower 
magnesium content is covered by a fairly thin mantle of Union Silt Loam. The 
dense oak-hickory upland forest on the Union Silt Loam gives way to red cedar- 
chinquapin oak forest at the periphery of the glade, but wherever the "cotton- 
rock" occurs it drains the area sufficiently that the red cedar-chinquapin oak forest 
is fairly well stabilized. In winter and spring, as a consequence of greater precip- 
itation and less evaporation, the open glade ("cotton-rock" stratum) is wet, often 
with standing water, while in summer and fall, when precipitation effectiveness 
is least, the porosity of the rock permits rapid drainage which results in rather 
extreme dryness in the upper few feet of the stratum. I have been frequently 
observing several of these glades since 1947, and, besides the rather obvious seasonal 
cycle of wetness and dryness, there are fluctuations every few years which markedly 
affect the population density of the species and the cover. In order for the peren- 
nials, in particular, to be successful in the glade community, they must be able to 
withstand fairly great fluctuations of the environmental conditions. 

The Old Field. — The Old Field is situated on a gently sloping knoll below the 
orchid greenhouse. The upward slope (5° 10') is NNW facing SSE. The knoll is 
covered by Union Silt Loam and probably will eventually support an oak-hickory 
forest. On this knoll a plot 60,000 square feet in extent was laid out as a rectangle 
300 X 200 feet. 

Prior to 1925 the area was under cultivation in the traditional "Ozark fashion." 
A few acres of beans, peas, and corn were planted, rows running uphill, and tilled 
by means of the mouldboard plow and mule driven by a tenant farmer. In 1925, 
when the property was acquired by the Arboretum, this field was badly eroded and 
almost bare of vegetation, but since 1925 it has not been grazed or burned, and the 
revegetation has been accomplished in the last 28 years. 

At present, the arborescent growth on the Old Field consists of scattered juni- 
pers, black oak, white ash, and thickets of red elm. Wisteria is an abundant shrub. 



having invaded the field from plantings nearby. The upper edge of the sloping 
field supports a growth of sugar maple, white ash, black oak, shingle oak, post oak, 
black cherry, smooth sumac, buckbrush, and poison ivy. The middle portion of 
the field consists of thickets or clumps of red elm saplings with a maximum height 
of about ten feet, and a thick growth of smooth sumac and wisteria. On the lower 
slope which drains into a tiny creek are growing black oak, white ash, northern 



The oldest arborescent growth is made up of persimmon, post oak, black oak, 
and shingle oak, which apparently invaded the field approximately five years after 
its abandonment. Then northern red oak, sugar maple, white ash and red cedar 
came in six to eight years after abandonment. At any rate, the older specimens of 
red cedar were twenty years old, eight years having elapsed since abandonment of 
the field. The rate of invasion by red cedar depends on the balance of several 
variables such as the number and proximity of seed-producing plants, condition of 
the invaded area, amount of cover, slope, pH, etc. Drew (1942), in his studies of 
abandoned crop-land in the Cedar Creek area, Boone and Callaway counties, Mis- 
souri, found that five years after abandonment sassafras, persimmon, shagbark 
hickory, post oak, and shingle oak invaded; after six to seven years, American elm; 
after eight to ten, northern red oak, sugar maple, white ash, honey locust, and black 
walnut; after eleven to twelve years, red cedar and black oak. 

In the herb layer the general cover consisted of fairly evenly distributed clumps 
of broom-sedge with Kentucky blue-grass in swales and level places bordered by 
clumps of little blue-stem. On the up-slope, broom-sedge, tall red-top, Canada 
goldenrod, and old field goldenrod were conspicuous. From midway to the bottom 
of the slope were these same species plus dewberry (Rubus flagellaris) and Aster 
pilosus. Here and there were almost bare sandy slopes, a few feet in extent, con- 
sisting of stands of triple awn-grass with a few lichens and mosses. Directly 
beneath the crowns of most of the larger trees were areas covered by lichens and 


Cedar Hill. — Another old field, here called Cedar Hill, is an exclosure with a 
gentle slope from west to east mowed annually with a sickle-bar and covered with 
well-spaced red cedar, black oak, black cherry, honey locust, and red elm. Th 
shrub layer consists of dewberry (Rubus flagellaris), smooth sumac, trumpet-vine, 
buckthorn, aromatic sumac, large clumps of buckbrush, partridge pea, and seedlings 
of sassafras, shingle oak, and persimmon. The herb layer consists principally of 
Canada blue-grass, triple awn-grass, tall red-top, Panicum capillare, and frequent 
clumps of broom-sedge, scattered clumps of little blue-stem, and occasional clumps 
of mountain mint and ironweed. Dewberry (Rubus flagellaris) was scattered 
fairly evenly throughout the field. 

The junipers average about 30 feet in height and thirty years of age, somewhat 



[Vol. 42 


negligible. Although there were abundant seedlings, particularly at the edge of 
the crowns of those mature junipers bearing berry-cones, all were below sickle-bar 
height and usually under three years of age. Just outside the exclosure and in the 
general area juniper reproduction was very good, and ages up to thirty-five years 
were represented. Species composition was not analyzed because of the artificial 
situation. In the meadow the juniper population in an area of 60,000 square feet 
was sampled. In general features this meadow is closely similar to the Old Field. 
The soil is Union Silt Loam but much deeper since Cedar Hill lies above an old river 
channel. The Cedar Hill population was included in this study because of its 
intermediate position morphologically between the Glade junipers and those of the 

Old Field. 


Several factors were studied for the Glade and Old Field in order to make com- 
parisons. Species composition, distribution of juniper seedlings, and the population 
structure of junipers were studied, and the data appear in Tables I— II, figs. 1—3, 
and pis. 20—21. Frequency was used as a quick method of measuring species com- 
position. The Glade and the Old Field were divided into four plots by staking out 
the diagonals. Along the diagonals alternating from one side to the other, 20 
quadrats (1 meter in size) were laid out, equally spaced from the center. Prelim- 
inary work was done in late spring of 1952, but other surveys were made in late 
spring and late summer of 1953. The presence of species in each quadrat was 
tabulated and the results for the Glade and Old Field appear in Table I. The areas 
studied seemed sufficiently homogeneous for the use of 1 -meter quadrat size, specif- 
ically to determine the species composition of the herb layer. A sample area about 
0.5 per cent of the total was used. A compromise in favor of a small sample area 
was made because of the limited time the class could be in the field. 

Variations of the juniper populations were measured by pictorialized scatter 
diagrams as demonstrated in Juniperus in a previous paper (Hall, 1952). Five 
characters were scored: gland length- width ratio; length of whip (longshoot) 
leaf; length of terminal whip; length of lateral whip; on secondary shoots, the per 
cent of tertiary branches which have the decussate leaf arrangement. Two of 
these characters were also analyzed by use of the parameters of the log-normal 
curve which the data fit. These data, when plotted on logarithmic probability 
paper, result in a straight line. 

In an attempt to get some estimate of seed-dispersal pattern, the distribution of 
seedlings was studied. Observations were made to determine the role that birds, 
mammals, and gravity may play in seed dispersal. 


Rough comparisons were made of species composition of vascular plants for the 
Glade and the Old Field. Only one quadrat size was used and no attempt was 
made t o obtain quantitative data on density or cover. For convenience a frequence 
index after the method of Raunkiaer (1918) was made up for the Glade and the 



Old Field, The percentage of frequence for a given species is the percentage ratio 
the plots on which the species occur bear to the whole number of plots sampled. 
Raunkiaer found that the greatest number of species had the least frequence; and 
that as the frequence increased the number declined steadily, until at the highest 
frequence the number increased slightly. Raunkiaer used five frequence groups, 
A, B, C, D, and E, designating the species of frequence 1-20 per cent, 21-40 per 
cent, 41-60 per cent, 61-80 per cent, 81-100 per cent, respectively. His law of dis- 
tribution, determined from quadrat studies in Europe was A>B>C — D<E; 

the actual percentages were approximately 53, 14, 9, 8, 16. In eastern America, 
Kenoyer (1927), found that 69, 12, 6, 4, 9 per cent were more characteristic. 

In this study the Glade had a frequence index of 28, 8, 2, 2, 2 or 
A>B>C = D : =E. This index is similar to that found by Kenoyer in his 
studies for America and indicates a fairly high proportion of sporadic species and a 
very few acting in a "dominant" role. The Glade has uniformly distributed patches 
of Sporobolus neglectus, and the size of each patch fluctuates somewhat from year 
to year. When the patches are smaller, more Bouteloua curtipenJula and forbs are 
in evidence. In good years the Glade has a good cover contributed principally by a 
very small number of species even though there is an appreciable number of 
sporadic species between the clumps of the cover plants. 

The Old Field index was 22, 3, 8, 5, or A>B<C>D>E. This somewhat 
irregular result is a product of the heterogeneous nature of the Old Field flora. 
There are a large number of sporadic species, a fairly large number of species inter- 
mediate in frequence, and no "dominants." The floral composition and vegeta- 
tional structure are markedly different from those of the Glade. Factors of 
competition and selection are less delicately balanced in the Old Field, and these 
factors undoubtedly cause the population structure of the junipers there to be 
different from that of the Glade. 

Table I lists the species and their frequencies found in sampling 0.5 per cent 
of the area studied. Only the under-story is represented. Woody plants in the 
list were seedlings or sprouts in size class 1 (0-0.9 feet) of Weaver and Clements 
(1938). The Old Field had seven species of woody plants represented in this size 
class for the area surveyed, which compares with nine species of similar size found 
by Drew (1942) to be common in old fields twenty-five years after abandonment. 
In the Old Field all sizes of woody plants were abundant up to size class 4 (3.6-9.5 
inches D.B.H.), but on the Glade very few specimens larger than size class 1 were 
seen, suggesting a high seedling mortality. The frequency data in Table I show 
that the Glade and Old Field are quite different in species composition and suggest 
a difference in the pattern of succession. The Glade is a "prairie" association with 
Andropogon scoparius and Rudbeckia missouriensis contributing most to its aspect 
and with Andropogon scoparius and Sporobolus neglectus contributing most to the 
cover. Even though there is some fluctuation in abundance from year to year, 
there are definite dominants, a good variety to the flora, and a stability in its over- 
all composition which suggests that succession is very slow and that the Glade 

[Vol. 42 

















3 6. 

Sporobolus neglectus 
Car ex Crawei 
Andropogon scoparius 
Houstonia langifolia 


Euphorbia corollata 
Rudbeckia missouriensis 

Sisyrinchium cam pest re 

Hypericum prolificum 
Gaillardia pule hell a 
Echinacea pallida 
Kuellia humilis 
Oenothera missouriensis 
Heliotropium tenellum 

Viola pedata 
Selaginella rupestris 
Juniper us virgin tan a 
Viola papilionacea 

Coreopsis lanceolata var. villosa 
JJlmus rubra 


Eleocharis camp res sa 

Aster oblongifolius 

Zizea aurea 
Monarda citriodora 

Cercis canadensis 

Rhamnus caroliniana var. 

Comandra Richardsiana 

Lotus americanus 

Orobanche uni flora 

Acalypha graciliens 

Panicum virgatum 

Triodia flava 


Bumelia lanuginosa var. 


Fraxinus am eric ana 

Monarda Russeliana 

Acer sac char um 

Anemonella thalictroides 

Bouteloua curtipendula 

Silphium laciniatum 

Psoralea esculenta 

Quercus Muehlenbergii 




.3 5 




Old Field 

Poa pratensis 
Rubus flagellar is 
Andropogon virginicus 
Aster pilosus 
Desmodium glabellum 

Rumex hastatulus 
Oxalis stricta 
Panicum virgatum 

Erigeron strigosus 
Taraxacum officinale 
Hieracium Gronovii 
Campsis radicans 
Triodia flava 

Achillea millefolium 
Parthenocissus quinque folia 

Ulmus rubra 

Celtis occidentals var. pumila 

Hypericum punctatum 

Prunus serotina 

Cercis canadensis 

Rhus aromatic a 

Juniperus virginiana 

Asplenium pi aty neuron 

Ambrosia artemisifolia 

Allium vineale 

Solanum carolinense 

Carex Haydenii 

Eragrostis spectabilis var. sparsihirsuta 

Veronica arvensis 
Asclepias syriaca 
Acer saccharum 

Phy satis pruinosa 
Pentstemon pallidus 
Sporobolus neglectus 
Daucus pusillus 
Antennaria fallax 
Potentilla recta 
Galium aparine 
Poa campressa 






Old Field 

56. Pellaea atropurpurea 

57. Vimbristylis Drummondii 
5 8. Petalostemum purpuretim 

59. Symphoricarpos orbiculatus 

60. Pentstemon pallidus 


Rhus Toxicodendron 

Nomenclature after Gray's Manual of Botany, 8th edition, 1950. 

community is one of real integrity. Because of the uniformity in distribution of 
Andropogon, Sporobolus, and Bouteloua the quality of the cover on the Glade can 
be determined fairly well with frequency data. In a good year the clumps of 
Andropogon are larger, denser, and more frequent. The Sporobolus effectively 
fills in between the clumps. In a dry or unseasonable year the Andropogon may 
lose ground, the Sporobolus to a lesser extent, but the Bouteloua may increase con- 
siderably. Thus, one can tell when the Glade is in good condition, so far as climatic 
fluctuations are concerned, by the relative frequencies (or preferably cover) of 
Andropogon scoparius and Bouteloua curtipendula. I think the Glade's general 
aspect and its severe selection against trees, shrubs, and woodland herbs, particularly 
as evidenced by selection against all but very Ashei-like junipers, support the con- 
clusions of Erickson, Brenner, and Wraight (1942) that these glades are edaphic 


The frequency data show the Old Field to be fairly typical for the Ozarks. 
There are several abundant species, but none of these contribute decisively to the 
field's aspect or cover. Poa pratensis, Andropogon virginicus, Panicum virgatum, 
Triodia flava, and Rubus flagellars are conspicuous. While the number of species 
is much the same for the two areas, the Old Field has fewer species with high 
frequencies or very low frequencies, and more with average frequencies. The Old 
Field is obviously in a rapid stage of succession toward woodland. 

Several other species were not found in the quadrats sampled. Mostly, these 
are not common plants in these areas but in some cases may occur with frequencies 
of 5 or 1 per cent in localized areas of the Glade or Old Field. The following 
species occurred on the Glade but did not occur in the quadrats: Ulmus rubra, 
Celtis occidentalism Physocarpus opulifolius, Amelanchier arborea, Rosa Carolina, 
Prunus serotina, Rhus glabra, R. aromatica, Ilex decidua, Celastrus scandens, Acer 
saccharum, Cornus Drummondi, Diospyros virginiana var. pubescens, Fraxinus 
quadrangulata, Viburnum rufidulum, Sporobolus heterolepis, Elymus virginicus 
forma hirsutiglumis, Allium stellatum, Menispermum canadense, Psoralea psora- 
lioides var. eglandulosa, Linum sulcatum, Petalostemum purpureum, Croton 
monanthogynus, C. capitatus, Asclepias tuberosa ssp. interior, Triosteum perfoli- 
atum, Vernonia crinita, Kuhnia eupatorioides var. corymbulosa, Hypericum per- 
foratum, Amsonia illustris, Isanthus brachiatus, Ruellia humilis, Solidago rugosa 
var. aspera, S. rugosa var. celtidifolia, S. nemoralis, and Parthenium hispidum. A 

[Vol. 42 


total of 80 species of vascular plants was collected on the Glade (45 from the 
quadrats; 35 by systematic search). Eighteen other species, mostly seedlings of 
the woody species already mentioned, were found in the Old Field but not in the 
quadrats. According to experience and estimates, it seems likely that slightly over 
100 species may be found in similar areas the size of the Glade and Old Field. Since 
these studies did not run continuously through the seasons involved, a considerable 
number of species was probably missed. It is also obvious from the data that a 
sampling area 0.5 per cent of the total is too small for best results; an area between 
1 and 2 per cent would have been effective. 

>sed glade is a tension zone between the open glade and the oak-hickory 



most of the cover in the arborescent layer is from juniper and chestnut oak. The 
transition from one community to the other seems abrupt since it occurs over a 
short distance, usually from 50 to 100 feet. From transect studies in the closed 
glade I listed 75 species, 39 of which were frequent species in the oak-hickory 
woodland and 43 were frequent in the open glade (Glade). 

Plate 20 shows the aspect of the Glade with close-ups of two quadrats (1 meter 
in size) on the up-slope and two on the down-slope. Plate 21 shows the Old Field 
aspect with two quadrats up-slope and two down-slope. 


The diagonals and sides of the Old Field and the Glade were used as transects 
for seedling counts. All juniper seedlings were counted within one-half meter on 
either side of the diagonals and one meter to the inside of the area boundaries. 
Seedlings were divided into three classes as follows: Class 1, 18 inches or less; 
Class 2, 18.1-36 inches; Class 3, 36.1-72 inches. Total number of seedlings re- 
ported is based on counts in approximately one twentieth of the total area. 

In the Old Field, a total of 621 seedlings was estimated, of which 276 were in 
Class 1, 207 in Class 2, 138 in Class 3. The actual number is probably somewhat 
higher since the smaller seedlings are rather easily overlooked. The seedlings were 
not distributed at random, but were either clumped about the mature (fruiting) 
female trees (with over 90 per cent of them on the down-slope side) , or were 
more or less aggregated at the bottom of the slopes or in other areas where materials 
carried by run-off water are deposited. Birds play an important role in the distri- 
bution of seedlings in the Old Field. Many young seedlings were clustered below 
the branches of shrubs and trees other than junipers. A very few were found on 
isolated knolls and in the open where they may have been deposited by birds in 
flight or by ground-feeding birds. 

The Glade supported a great many more seedlings under four years of age than 
did the Old Field. The estimate was: Class 1, 6,672; Class 2, 10; Class 3, 12. 
The total estimate was 6,764 seedlings for the 60,000 square feet of glade exclosure. 
These were distributed much as in the Old Field, but even more uniformly because 
of the greater uniformity of the 10° slope characterizing the Glade. The seedlings 



were either clumped on the down-slope side of the mature (fruiting) female trees 
or were aggregated at the bottom of the Glade slope where the Sporobolus neglectus, 
because of its increasing cover toward the lower edge, acted as a brake to tumbling 
berry-cones carried by run-off water. There were numerous seedlings in the open, 
probably from seeds deposited by birds. Great flocks of robins and cedar waxwings 
are a common sight on the glades, particularly in the fall after the crop of berry- 
cones has ripened. Fortuitous fluctuations in climate probably add to the seedling 
mortality. In very wet seasons the seedlings with more genes of red cedar are 
probably favored only to be wiped out in a later year by drought. Likewise, 
seedlings possessing a strong /. Ashei component are favored in drought seasons 
but may suffer defeat in wetter periods. 


The variations in the junipers (Jtiniperus virginiana) of the Glade, Old Field, 
and Cedar Hill were studied, and five morphological characters were measured and 
plotted as pictorialized scatter diagrams. Two of the five characters, measured, 
gland length-width ratio and lateral whip length, were used on the ordinate and 
abscissa of logarithmic paper (1X1 cycle) and the other three as rays on the 
scatter diagrams. The data for the three characters plotted as rays were grouped 
as follows: (1) whip-leaf length: long-bar = 8 mm. or more, half-bar = 5-7 
mm., no bar = 2-4 mm.; (2) per cent decussate: long-bar = 25-100 per cent, 
half-bar = 6-24 per cent, no bar = 0-5 per cent; (3) terminal whip length: 
long-bar =80 mm. or more, half -bar =± 31-79 mm., no bar = 0-30 mm. The 
scatter diagrams for the Glade, Cedar Hill, and the Old Field populations, as well 
as for populations typical for the species, are shown in fig. 1. 

Population means and ranges were obtained from the grouped data (pictorialized 
scatter diagrams) in the manner described in an earlier paper (Hall, 1952, p. 53). 
The numbers given have no absolute value but represent a graded scale or index of 
characteristics. The scale is made up in such a way that low scores belong to 
/. Ashei, high scores to J. virginiana, and intermediate scores to morphological 
intermediates or introgressants of the two species. From these scatter diagrams 
the following population means on this arbitrary scale were obtained: Glade — 4.35; 
Cedar Hill — 5.48; Old Field — 10.0. Typical J. Ashei has a population mean of 
2.0-2.3; typical /. virginiana, Eastern race, has a mean of 9.0-10.0. 

The two characters plotted on the abscissa and the ordinate for the pictorialized 
scatter diagrams were also analyzed by using ordered values. The data were ex- 
pected to fit a log-normal distribution which was confirmed, for the data produced 
a straight-line function when the ordered values (from smallest to largest) were 
plotted on logarithmic probability paper. I followed the practical short-cut recom- 
mended by George W. Thomson of the Ethyl Corporation, which used the range 
as a measure of dispersion instead of the standard deviation. Statistical techniques 
using the range are particularly efficient with small samples. 

This is an excellent method of checking the confidence of data where only an 

average and the range are available. The confidence limits a, b = x ± ts/v n > 


[Vol. 42 



Juniptrut vlrqlnlono L. 



Juniptrut virgin! 


Juniptrut Athti 







i L 



tOO JOO 400 500 900 7C0 #00 *X> 



# nilMrtUk. *HlP LCNCTM 


too JOO *oo 


9O0 100 100 IO0 

Juniperus Virginia na L 









*o WO 




tOO JOO 400 


300 «00 TOO #00 900 

Juniptrut virginiano L. 



)t V 











, »0 100 


• %OtCU»tAT| UAVC* 



Fig. 1. Scatter diagrams showing the variations in five characters of a typical race of Juniperus 
virginiana from southwestern Virginia, a typical race of /. Ashes from the Edwards Plateau of 
Texas, and of three populations of /. virginiana at Gray Summit, Mo. The diagram at the upper 
left for the typical races shows the means (large dots), the extremes (small dots), the limits of 
variation (the curved lines). (The symbol in the space at the upper right of this diagram should 
have had three long arms.) In the diagrams for the three Gray Summit populations each dot rep- 
resents a single individual. 





lation variance, and n is the number in the sample. Or, by using the short-cut the 
95 per cent confidence limits are equal to the range multiplied by the proper value 1 
(0.1064 for n = 25, 0.0720 for n 



(log values) for two characters of each of the three populations and typical popu- 

/. Ashei and / 

Figure 2 represents the ordered values 

plotted on logarithmic probability paper. The slopes of the lines depend on the 
amount of variation within each group; the greater slopes indicate the more 
heterogeneous populations. The point of intersection of each line with the 50 per 




of the 25 per cent and 75 per cent values with each line delimit the range where 
half the values for that character will probably lie. In fig. 3 rectangles are made 
about the means to correspond to the 5 per cent probability level, and the farther 
two rectangles are separated the greater the probability that the difference between 

ling means is not due to chance. The dotted lines indicate the 95 
per cent confidence limits of the means from Table II. 

These five populations were compared to determine the probability that they 
represent the same population with equal means by means of the t-test. If one 
assumes the means of two populations to be equal, the sample difference is tested 
to see whether it is no more than sampling variation from the hypothetical dif- 
ference, zero. Each of these comparisons led to a large t value, indicating a low 
probability that they represent samples from a single population. Welch's modifi- 
cation of the t-test was used for the comparison because the variances were not 
homogeneous (Comparison of /»i— /i 2 regardless of <j x /<t 2 )• 





/. Ashei 
Cedar Hill 
Old Field 
/. virginiana 




Whip leaf length (mm.) 


Conf. limits I Stand." 


30 yrs. 



35 yrs. 



30 yrs. 



20 yrs. 



30 yrs. 








*Log values. 

Gland length/width 


Conf. limits 














See the Appendix to this paper. The values for these multiplying factors v 
sample numbers from 2 to 1000 by George Thomson and appear in Table III. 


[Vol. 42 







J. virginiana 




J. Ashei 



20 30 40 50 60 70 80 


90 95 














J. virginiana 






J. Ashei 










J I I 






20 30 40 50 €0 70 80 


90 95 


Fig. 2. Ordered values, plotted on logarithmic probability paper, for two char- 
acters for typical populations of Juniperus virginiana, /. Ashei, and the three popula- 
tions of /. virginiana from Gray Summit. Length of the lines delimit the range where 
95 per cent of the values for these two characters will probably lie. 































Fig. 3. Variation in whip leaf length and gland length-width ratio in the three 
populations of Juniperus virginiana from Gray Summit, Mo. The estimated means 
are represented by vertical bars; the 95 per cent confidence limits of the means by 
vertical dashes; the rectangles delimit the ranges where 50 per cent of the values will 
probably lie; the length of the horizontal lines delimit the ranges where 95 per cent 
of the values will probably lie. 


In the Missouri Botanical Garden Arboretum at Gray Summit, Missouri, these 
three populations of Juniperus virginiana called here Glade, Cedar Hill, and Old 
Field, are each distinct, having considerably different variances and significantly 
different means. Each is also distinct from either species, /. Ashe? or /. virginiana, 
which contribute to the variability of these Ozark populations. Yet, these three 
Ozark populations exist within less than a mile of one another, but in habitats 
which, at least in part, may be noted for their differences more than their similar- 
ities. In the sums of several morphological characters these populations differ from 
one another in nearly the same degree, so that a crude ranking using five characters 

[Vol. 42 



gives arbitrary values of 1.0 for J. Ashei, 2.1 for Glade, 3.0 for Cedar Hill, 4.0 for 

Old Field, 4.9 for /. virginiana. 

Where each character is compared separately there is greater apparent difference 
between populations. Methods utilizing several characters therefore give a more 
accurate picture of the population. With several characters plotted, scatter dia- 
grams give a good impression or "bird's eye view" of the variation within a popu- 
lation even though statistical inferences may not be drawn from them alone. When 
the populations are compared character for character, one finds that the degree of 
difference between characters is not necessarily similar for each character. The 
Glade population may closely resemble Cedar Hill in one character and more 
closely approach /. Ashei in another. Unequal variability, character for character, 
is characteristic of species hybrids and introgressant populations. Even when con- 
sidering variation within the individual, most naturalists know that hybrids, par- 
ticularly in wide crosses, have more intra-individual variability than do more typical 

plants of the species. 

The Glade and Cedar Hill populations, both of the Ozark race, i. e. intro- 

gressants of /. virginiana by /. Ashei, are the most variable. The Old Field popu- 
lation is a mixture of the Ozark race and the Northern race (including var. crebra 
type) of red cedar; but as far southwestward as Gray Summit, Missouri, it occurs 
sporadically in disturbed habitats where the soil is sandy, light, and well drained. 
I have not yet seen this northern invader on a glade even though it could con- 
ceivably occur there within a favorable micro-habitat. 

The Old Field and Cedar Hill habitats are very similar in every aspect but 
history of use by man. The Old Field was allowed to erode severely, and conse- 
quently it has less topsoil and organic matter than the Cedar Hill. The Old Field 
soil is more acid, sandy and coarser than Cedar Hill soil. Both Old Field and Cedar 
Hill soils are quite different from the very thin, organically rich, spring-wet, 
summer-dry soil of the Glade. 

The Northern race of red cedar (including var. crebra) tends to be arenicolous, 
and even though it develops well on sandy areas overlying limestone or on the 
strands of marly lakes, it does best on slightly acid, well-drained strand or dune 
areas where competition with other plants is at a minimum. The Northern race 
becomes more arenicolous as it becomes more columnar or spire-like. Because of 
the ease with which junipers may be distributed over great distances, it is not sur- 
prising to find an occasional mixture of the Northern race of red cedar in the 
northeastern Ozarks. However, its occurrence there is a result of man inad- 
vertently preparing a suitable habitat, an eroded, acidic, sandy field with poor 
cover. Except for the minor role the Northern race plays in introgression with 
the Ozark race, its effect on the Ozark populations is transient and dependent on 
the perpetuation of poor land-use practices. Ozark fields which have had fair 
treatment, particularly where cover has been protected, support stands of the 
Ozark race alone. In part, the Ozark race is also dependent on man since a large 
number of its populations occur on cleared or grazed land. 



Several other factors affect the distribution and structure of juniper popula- 
tions. Some of these factors are intrinsic and some extrinsic, but it is impossible 
to consider them independently. Rate of reproduction, per cent viability, presence 
of introgressive genes, character of the climatic extremes (whether unfavorable 
for natural seed stratification, which seems to be the situation at the northern limits 
of red cedar), amount of bare area in the habitat, slope, soil conditions, presence 
and abundance of bird and mammal species which aid in seed dissemination, role 
of run-off water and gravity in seed dissemination, and several other factors occur 
in different combinations and degrees to produce populations ranging from uniform 
(as in a cedar-brake) to less uniform (apparently random) to clumped. In gen- 
eral, when a large enough area is examined there may be found "parent" popula- 
tions consisting of cores of ancient specimens which supply or supplied the seed 
stock for the general area. In southwestern Missouri these mother-plants may be 
found on the bluffs of the White River where they have fruited for centuries safe 
from fire and other hazards. In northern Missouri the bluffs of the Missouri and 
Meramec Rivers support populations of ancients. These bluff habitats of the major 
rivers were probably the original environment for J tint perns long before the Ozark 
glades were first colonized by junipers. 

On most of the glades in the eastern Ozarks the oldest junipers are over 100 
years of age, probably between 125 and 135 years. Since all the older junipers 
from which increment borings were taken were hollow at the center, it was not 
possible to get precise ring counts, but by applying a factor determined by the 
rate of decrease in years per inch from the center to the outside of an increment, a 
good estimate of age may be made. Much older trees (judged by circumference) 
have been found occasionally on bluffs throughout the Ozarks, but those from 
which increments were taken were hollow and usually with more or less eccentric 
rings. Such old junipers are not found in Ozark fields. In the early years of the 
19th century and previously, written accounts of travels and casual reports by 
geologists indicate that junipers were to be found only along bluffs of the major 
rivers, but reports in the last half of that century presented a picture of juniper 
distribution much as it is today. Wherever the rocky limestone glades are burned 
each year, they are devoid of junipers and shrubs, giving an open grassland aspect 
with an abrupt transition to forest at the edges. Also, the junipers do not easily 
invade a glade that has been protected from grazing and supports a good stand of 
grass. The treeless glades, which are not uncommon, always have good cover and 
protection from man's grazing animals. Where glades show signs of erosion 
junipers are localized. In general, juniper density is proportional to the degree of 
land abuse, so that glades may support junipers distributed as dense "brakes", open 
stands with evenly but widely spaced individuals, or scattered, clumped colonies. 
Because of the nature of the glades slight grazing could easily weaken the cover to 
permit invasion of junipers where fire-sere grassland would not be so easily weak- 
ened. Before white man came to the Ozarks, wild herds probably initiated the 

[Vol. 42 



invasion of the glades by junipers; white man tremendously accelerated the process 

through his domestic animals. 

In an earlier paper (Hall, 1952) I suggested that introgression between /. Asbei 
and /. virginiana may have begun as early as the late glacial period, followed by an 
expansion during the xerothermic period. Throughout this time, before settlement 
of the Ozarks, the most continuous habitats for junipers were the bluffs, knobs, 
rough glades, and youthful streams or rivers with flood-plains in an early stage of 
succession. Probably continuously the bluffs, knobs, and glades (edaphic sub- 
climaxes) were inhabited by Ashei-Yike junipers while the streams and river banks 
supported scattered stands of virgin:ana-\ike plants. 

The distribution pattern of junipers in the Ozarks has probably changed 
radically since settlement. It is clear that the oldest junipers are situated on bluffs 
or glades, while old fields in the St. Louis area may support mature colonies (oldest 
plants up to 70 years) of the Ozark race or occasionally more youthful colonies 
(40 years or less) of the Northern race mixed with Ozark. From this study and 
cursory examination of other fields, it is evident that the condition of the land in 
large part determines the kind of junipers selected from the local gene pool. 
Literally, explosive distribution of juniper has occurred in the last hundred years, 
progressively increasing in amplitude as abused or worn-out land was abandoned 
to old-field succession. The fact that these neighbor populations, particularly 
Glade and Old Field, are distinct even though intermediate does not mean that 
hybridization between them is limited, but that each habitat is selecting the best- 
adapted genotypes. Natural selection is strong enough to limit the portion of the 
available gene pool which may be realized. Here are two powerful forces working 
against one another. Strong natural selection restricts the gene pool; hybridization 
amplifies it. With species like those of Juniperus, which hybridize freely and are 
readily dispersed, I suggest that continuously variable habitats result in continu- 
ously variable populations, and the simplest situation is when the hybridization is 
between two allopatric species. 

Anderson (1948) stated that hybrids between two species differing in habitat 
requirements are expected to occupy intermediate habitats. The F 1 should occupy 
a habitat more or less intermediate between the two species; the F 2 should occupy 
more variable habitats, but the range should lie between the habitats of the species; 
backcrosses or introgressants should occupy habitats closest to the backcross parent. 
Two species may be quite interfertile, exist side by side, yet show little evidence of 
hybridization between them because the habitats select the parent-type combina- 
tions. If the habitat is disturbed, these available mongrels may become established. 
The homes of these variants are the seres either initiated by natural processes or 
man. While hybrid swarms and hybrid habitats are transient, they have their in- 
fluence toward increasing variability both through cryptic processes, e. g. structural 
differentiation, and toward accelerating regional differentiation. If, for the sake 
of discussion, man as a producer of hybrid habitats is ignored, it will be seen that 
processes which initiate succession produce hybrid habitats, and indeed the edges 



of ecologic systems, physiographic provinces, and climatic regions are hybrid 
habitats. Ecologic tension zones are hybrid habitats where species which meet may 
develop discordant variability. When man enters the scene, the pattern becomes 
more complex. Hybridization may do its cryptic work in tension zones through 
all levels of ecologic structure. The obvious evidence of hybridization, the hybrid 
swarms, are transient; and they may occur whenever the right hybrid habitat is 
available, but in most situations the habitats select gene combinations more like 
the parents and tend to keep the populations distinct. 


In the northeastern Ozarks Juniperus virginiana is represented by the Ozark 
race (introgressants from /. Ashei) on bluffs, glades, and most old fields. Oc- 
casionally, the Northern race is found locally with a little admixture of the Ozark 
race. Three populations of /. virginiana in the northeastern Ozarks were studied 
in detail in order to compare their variation patterns and habitats. Variation 
patterns were compared by means of scatter diagrams for five characters and 
ordered values, means, 95 per cent confidence limits, and standard deviations for 
two characters. The t-test modified by Welch was used to compare the means of 
the populations. Habitats were compared in terms of their histories, slope, 
soil profile, flora, frequency of species, rough estimates of cover, seedling 

The evidence suggests that the three populations — Glade, Cedar Hill, and Old 
Field — are distinct and differ more or less proportionately to the differences in their 
habitats. The Glade (Ozark race) is the most southwestern in affinity, more 
closely resembling a population and habitat of /. Ashei. Cedar Hill (Ozark race) 
is intermediate between Glade and typical red cedar and occurs on old fields which 
are in good condition or, farther southwestward from Gray Summit, Missouri, on 
more worn-out lands. The Old Field (Northern race with a little mixing from 
the Ozark race) occurs on worn-out acidic and sandy lands in the vicinity of St. 
Louis and northeastward. 

Extremely high seedling mortality (much higher on the Glade than Old Field) 
suggests that strong natural selection restricts the field for variability even in the 
face of considerable hybridization. The present balance between these two evolu- 
tionary forces depends on man making and keeping available a variety of suitable 
habitats. A shift in habitat qualities will result in a shift in the variation pattern 
of the junipers. 

Distribution of age classes in the junipers indicates that the bluffs, knobs, and 
glades have been colonized longest by junipers, followed by the old fields supporting 
the Ozark race, and last the worn-out sandy, acidic fields supporting youthful 
colonies of the Northern race. An explosive expansion of juniper colonization 
resulting from man's activities seems to have occurred within the last hundred 
years, growing progressively as land has been worn out and abandoned. It is 
proposed that man is primarily responsible for the present explosive evolution in 
Juniperus in eastern North America and that if his influence were removed from 

[Vol. 42, 1955 J 


the scene, the field of variability would decline as the area of occupation contracts. 
Without man's help junipers would be more or less restricted to the areas in a 
natural state of arrested development such as the bluffs, knobs, and glades. 


Anderson, Edgar (1948). Hybridization of the habitat. Evolution 2:1-9. 

Beilmann, August P., and Louis G. Brenner (1951). The recent intrusion of forests in the Ozarks. 

Ann. Mo. Bot. Gard. 3 8:261-282. 
, , (1951). The changing forest flora of the Ozarks. Ibid. 38:283-291. 

Braun, E. Lucy (1950). Deciduous forests in eastern North America. Philadelphia. 

Drew, William B. (1942). The revegetation of abandoned crop-land in the Cedar Creek area, 

Boone and Callaway counties, Missouri. Univ. Mo. Coll. Agr., Agr. Exp. Sta. Res. Bull. 344. 
Erickson, Ralph O., Louis G. Brenner, and Joseph Wraight (1942). Dolomitic glades of east-central 

Missouri. Ann. Mo. Bot. Gard. 29:89-101. 
Etter, Alfred G. (1953). Wildwood — A study in historical ecology. Ibid. 40:227-257. 
Fernald, M. L. (1950). Gray's Manual of Botany. Eighth Edition. New York. 
Hail, Marion T. (1952). Variation and hybridization in Juniperus. Ann. Mo. Bot. Gard. 39:1—64. 
Kenoyer, Leslie A. (1927). A study of Raunkaier's [Raunkiaer's] law of frequence. Ecology 

Raunkiaer, C. (1918). Recherches statistiques sur les formations vegetales. Kgl. Danske Vidensk. 

Selskab. Biol. Meddel. l 3 :l-80. 
Sauer, Carl O. (1950). Grassland climax, fire, and man. Jour. Range Management 3:16—21. 
Steyermark, Julian A. (1940). Studies of the vegetation of Missouri. I. Natural plant associations 

and succession in the Ozarks of Missouri. Field Mus. Nat. Hist. Bot. Ser. 9:349—475. 
Thomson, George W. Personal communication. (Also, see Appendix). 
Thornthwaite, C. "W. (1931). The climates of North America according to a new classification. 

Geographical Review 21:63 3—655. 
Weaver, John E., and Frederic C. Clements (1938). Plant ecology. New York. 

Explanation of Plate 


Aspect of Glade 

Fig. 1. Open glade looking south. 

Fig. 2. Open glade looking north. 

Figs. 3 and 4. Two quadrats in open-glade up-slope. 

Figs. 5 and 6. Two quadrats in opcn-glade down-slope. 

Ann. Mo. Bot. Gard., Vol. 42, 195 5 

Plate 20 

■ ^ ■ 






Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 21 








[Vol. 42, 1955] 


Explanation of Plate 


Aspect of Old Field 

Fig. 1. Old Field begins beyond mowed area 

Fig. 2. Old Field, interior. 

Figs. 3 and 4. Two quadrats, up-slope. 

Figs. 5 and 6, Two quadrats, down-slope. 

[Vol. 42 



estimation of confidence limits of population mean from range 

I am very grateful to George W. Thomson for the method by which 95 per 
cent confidence limits of the mean may be rapidly estimated by multiplying the 
sample range by a factor. Lord (1947) prepared a table of factors for sample 
sizes from two to twenty for six classes of probability (90-99.9 per cent). Thom- 
son prepared a table of factors including greater sample sizes for 95 per cent 
probability. Thomson's table and brief discussion are published here for the benefit 
of other biologists who may wish to use this short method of getting 95 per cent 
confidence limits of the mean. 

The usual method of estimating the variability of a population from a sample 
is by the use of the usual standard deviation estimator, s. However, recent research 
has shown that much more convenient estimates can be obtained from the sample 
range, which is defined to be the difference between the smallest and the largest 
values in the sample. These estimates are not as efficient in the statistical sense, but 
the loss is not important from a practical point of view. A useful by-product of 
this research is the rapid estimation of confidence limits of the mean of a normal 
population by the multiplication of the sample range by a factor. These results 
are associated with a smaller number of equivalent degrees of freedom than con- 
fidence limits based on the usual s estimates of the population standard deviation. 
A close approximation for the multiplying factors was found by Patnaik (1950) 
who approximated the distribution of the range in normal samples by the x 

If w is the range in a sample of size n, then w/cr 

c x /V 

v w 


c is a 


factor and v is an equivalent number of degrees of freedom for x* 

The following table extends previous calculations (Thomson, 1953) to sample 
sizes as large as 1000. 





Numb, in 

Degrees of 







Confidence limits** 





5% t 

± t/(cVn) 














































11 1 8.3485 










TABLE III (Continued) 










Numb, in 

Degrees of 



^^^fc £"^ m 4 # A * a 





Confidence limits** 




5% t 

± t/(cVtT) 










































































































































































* All quantities are believed to be correct to within 3 units in the last place shown. Values 
for n ss 2(1)10 are from Thomson (1953). 

** Multipliers for range to get 95% confidence limits of normal population mean. 

The following example demonstrates the use of the method. A sample size of 
twenty-five log values of whip leaf length from the Cedar Hill population of the 

[Vol. 42, 1955] 


Ozark race of Juniperus virginiana had a range of 0.637 and a mean of 0.842. It 
is estimated with 95 per cent confidence that the population mean lies within 
0.842 ± 0.067 or 0.775 to 0.909 (5.96 mm. to 8.11 mm.). 


Lord, E. (1947). The use of range in place of standard deviation in the t-test. Biometrika 34:41-67. 
Patnaik, P. B. (1950). The use of mean range as an estimator of variance in statistical tests. 

Ibid. 37:78-87. 
Thomson, George W. (1953). Scale factors and degrees of freedom for small sample sizes for 

^-approximation to the range. Ibid. 40:449-450. 



Emeritus Director 
George T. Moore 

Edgar Anderson 

Assistant Director 
Hugh C Cutler 

Carrozx W. Dodge, 


Robert E. Woodson, Jr., 

Curator of the Herbarium 

Henry N. Andrews, 


Roixa M. Tryon, 

Assistant Curator of the 

Hugh C. Cutler, 

Curator Museum of 
Economic Plants 

Gerald Ulrici 

Business Manager 

George B. Van Schaack, 

Honorary Curator of Grasses 

Julian A. Steyermark, 

Honorary Research Associate 

Frederick G Meyer, 


Alice F. Tryon, 

Research Associate 

John D. Dwyer, 

Research Associate 

Nell C Horner, 

Librarian and Editor 
of Publications 


John S. Lehmann 

Daniel K. Catlin 

Second Vice-President 
Eugene Pettus 

Leicester B. Faust 
Dudley French 
Henry Hitchcock 

Richard J. Lockwood 
A* Vessel Shapleigh 

Ethan A. H. S: 

Robert Brookings Smith 


Arthur C. Lichtenberger, 

Bishop i the Diocese c 

Stratford Lee \1orton, 

: the Academy £ 


t T 

Ethan A. H. Shepley, 


of Washington 

Mr Irma H Fried 

President c kar 

R/ \iond R. Tucker 

May the City 

Gerald I sti< Secretary 

Volume XLII 

Number 3 


of the 

Missouri Botanical 



Tassel Modifications in Zea Mays 

Norton H. Nickerson and Ernest 

Dale 195-212 

A New Species of Doryopteris from Surinan 

Karl I Kramer and Rolla M, Tryon, fr. 

I 3 


A Revision of the Genus Ce/astr 

Ding Hon 215-3C2 





M 1 


<nwl as *ecood-eiasi matt? 

s* A 

(► r 

fffcfl a :■ 



of the 

Missouri Botanical Garden 

Quarterly Journal containing Scientific Contributions from the 

Missouri Botanical Garden and the Henry Shaw School of Botany of 
Washington University in affiliation with the Missouri Botanical 

The Annals of the Missouri Botanical Garden appears four „ 

during the calendar year: February, May, September, and November. Four 
numbers constitute a volume. 

Subscription Price $10.00 per volume 

Single Numbers 2.50 each 

Contents of previous issues of the Annals of the Missouri Botanical 
Garden are listed in the Agricultural Index, published by the H. W. Wilson 


of the 

Missouri Botanical Garden 

Vol. 42 SEPTEMBER, 1955 No. 3 



Department of Botany, University of Massachusetts, Amherst and 

Kerckhoff Laboratories of Biology, California Institute of Technology, Pasadena 

Emeritus Professor of Biology, Union College, Schenectady, New York 

Zea Mays L., in addition to its great economic importance, is preeminent among 
plants as a tool in the study of heredity. However, even though our knowledge 
of maize genetics has made rapid and often spectacular advances, certain problems 
involving basic morphology of the plant have been left unanswered. Foremost 
among these is that of the pistillate inflorescence, or ear, the structure of which 
has been discussed elsewhere (Nickerson, 1954). Another equally basic problem 
involves the staminate inflorescence, or tassel. Anderson (1951) briefly sum- 
marized botanical knowledge dealing with this inflorescence, and Alava (1952) 
showed how certain maize races could be characterized on the basis of information 
obtained from their tassels. 

Preliminary studies on tassel morphology by the senior author indicate that 
the somewhat stereotyped and simple construction of tassels is misleading; certain 
parts apparently have been subjected to reductions even greater than those affect- 
ing corresponding ear parts. Morphological analysis of the tassel as it occurs atop 
a typical maize plant is even more difficult than morphological analysis of the ear. 

One source of information on ear morphology was through studies of certain 
genetic forms in which particular parts were present in an accentuated form. It 
was felt that the same technique might profitably be applied to the tassel, since 
variation in maize tassels is, as in other parts of this specialized grass, variation on a 
theme. Specimen tassels from the collecton of sixteen tassel mutants assembled by 
the junior author were therefore examined. They contained a wealth of material; 

1 Massachusetts Agricultural Experiment Station Contribution 994. This work, which was per- 
formed at the California Institute of Technology, Pasadena, was supported by funds supplied m part 
by the Atomic Energy Commission. The authors are indebted to Dr. E. G. Anderson both for his 
freely shared experience in careful field observation and for his critical reading of the manuscript. 
In addition, they wish to thank Dr. Edgar Anderson for his cooperation in this project. 

2 Present address: Division of Biology, California Institute of Technology, Pasadena, Calif. 


[Vol. 42 


a search of the literature revealed that, in addition to the fact that several of them 
had not been described, many of the known forms had very meager descriptions. 
Since knowledge of the tassel mutants in themselves will serve both as a prelim- 
inary step in analysis of tassel morphology and as a basis for their employment in 
studies involving gene interaction, fuller descriptions of these anomalous forms 
are given. 

"We have attempted to give below a careful description of how each mutant 
type affects structures present in tassels and ears of North American corn-belt 
maize. Plants of each mutant have been bred to various combinations of two 
widely known corn-belt inbreds, CC5 and L317, sufficiently often to allow mean- 
ingful comparisons to be made between them. It must be remembered, however, 
that the standard stock is largely responsible for these particular manifestations, 
and that the background here employed was one found to be well adapted to grow- 
ing conditions in southern California. Certain forms which might logically be 
included in this discussion, such as Corn-grass, Teopod, Silky- 1 and Silky-2, were 
not studied. Silkless, a well-known ear anomaly, also was not included, because in 
our stocks Silkless tassels are indistinguishable from those of Standard plants. 

It is possible to construct a key by which these tassel mutants may be sep- 
arated. However, since even in relatively homogeneous stocks, an example of one 
mutant may sometimes match the description of another, such an artificial device 
cannot be too heavily relied upon. A key is here intended only to show certain 
general trends among the sixteen forms; it may or may not indicate underlyin 
genetic and physiological similarities. The several descriptions given later are 
arranged in the numerical order indicated in the key. 


A. Tassel of normal (Standard) or nearly normal proportions 
B. Tassel with silks 

C. Pollen shed; fruits formed 

D. Fruits on proximal branch areas from imperfect florets Ts s — 6 

D. All fruits from perfect florets /$ z 8 

C. No pollen shed; no fruits formed /5 8 — 9 

B. Tassel without silks 

C. Glumes undeveloped „ y> 13 

C. Glumes well developed 

D. Glumes long (average 11.5 mm.) Tu — 14 

D. Glumes near normal length (average 8-10 mm.) 

E. No pollen shed „ m5l — 10 

E. Pollen shed 

F. Up to l / 4 or y$ of branch tips sterile; pedicellate spikelets often with several 

sets of florets bd — 11 

F. Branch tips never sterile; pedicellate spikelets not branched Club — 12 

A. Tassel of abnormal proportions 
B. Tassel with silks 

C. Half or more of tassels pistillate; staminate spikelets confined to middle areas or 
outer halves of branches; spikelets with no more than two florets each 

D. Branches indeterminate, mostly ending in immature pistillate structures Ts 3 — 3 

D. Branches otherwise 

E. Spikelets loosely spaced on branches f S2 — 2 

E. Spikelets crowded on branches ts± — 1 

C. Tassels either staminate, pistillate, or mixed; if mixed, staminate spikelets often 
with more than two florets 




modifications in Zfti May 


D. Extra florets developed on both sessile and pedicellate spikelets; average peduncle 

length 10 cm. or less Tst — 7 

D. Extra florets developed on pedicellate spikelets only; average peduncle length more 
than 10 cm. 

E. Seed set very sparsely if at all; average glume length 6 mm to — 4 

E. Seed set in noticeable quantities; average glume length 8.5 mm to* — 5 

B. Tassel without silks 

C. Branches horizontal or slightly lax with reference to central culm ra\ — 15 

C. Branches upright, close to central culm tai — 16 

Certain measurements have been made on at least fifteen and often twenty-five 
plants of each mutant form, and averages of these data are given in Table I. 
Peduncle length is the distance from the uppermost leaf-bearing node to the node 
at which the lowermost tassel branch arises. Tassel length refers to distance from 
the node at which the lowermost tassel branch arises to the tip of the central culm. 
Branching area is that distance along the central culm from the lowermost branch 
node below the central spike; it is included within the tassel length. Primary 
branches are axes of the second order, when the main culm of the plant is con- 
sidered to be an axis of the first order. The percentage values for peduncles and 
branching areas were obtained by dividing these respective lengths by the sum of 

peduncle and tassel lengths. 

These same data are presented graphically in fig. 1 by means of a pictorialized 
scatter diagram (Anderson, 1949). The diagram is meant only to indicate the 
average extent to which each of the mutants departs from average measurements of 
Standards. Limits were chosen so that Standard averages would always be char- 

























1 (to) 

2 (to) 

3 (Ts 3 ) 

4 (to) 
4 a (T5 4 a ) 


(Ts 5 ) 



Male-sterile 1 
Branched-silkless (bd) 

Club (club) 
Vestigial glume (Vg) 
Tunicate (Tu) 
Ramosa 1 (rai) 
Ramosa 2 (ra 2 ) 



-a !? 




o £ 

to U 














































O Xi 



6 £ 


J! v 

G rt 

3 « 





g" 5 

M * 





[Vol. 42 

acterized by long flags. Thus, similarity to Standard in both graph position and 
flag lengths means that a particular mutant departs relatively slightly from a 
normal form. The converse is also true; the farther away in position and the 
greater the flag deviations from Standard, the more a given mutant differs in form. 










is 4 a 






























K.r ~ L t Pkt ° rialized « ct V er di3gram ° f "«»**» contained in Table I, showing relationships 



Tea Mays 199 



Standards, L317 and CC5 (pi. 22, fig. 1), 

Various mixtures of these two corn-belt inbreds were used. The history of 
Lancaster Surecropper, from which M. T. Jenkins developed L317, has been re- 
viewed by Anderson (1944b). Wisconsin Inbred CC5 has an ancestry in the 
open-pollinated maize which was common to that section. Anderson and Brown 
(1952a, 1952b) showed that these corn-belt forms were various mixtures of 
Northern Flint and Southern Dent races. Nickerson further pointed out (1953) 
that these two races had previously been mixed in some degree, and that the 
excellent hybrid vigor manifest in present-day crosses, such as those between these 
two standards, is based on rather small differences. Ears are 14- to 18 -rowed, 8 to 
10 inches long, and bear bright yellow, dented kernels. 

1. Tassel-seed 1 (7s J (pi. 23, fig. 1), 

First described as "tassel-ear" by Emerson (1920), this mutant was classified 
as Tassel-seed 1 by Emerson et al (1935), and assigned by them to Chromosome 2. 
Emerson originally described the inflorescence as being completely pistillate, com- 
pact and distinctly ear-like, with "glumes and palae short, broad and rounded, in 
all respects much like those of true ears." He noted that these characters could 
be observed in both mature and immature tassels as well as in poorly pollinated 
mature ones. It may be added that each pair of pistillate spikelets is associated 
with a well-developed cupule (Nickerson, 1954). The branches are thickened so 
that they appear as wide green bands in adaxial view. In our cultures, as in Emer- 
son's, seed set was good and very often both florets of a spikelet develop a 
caryopsis. This development obscures any evidence of regular rowing in the 
central spike. 

Emerson noted the general plant weakness in comparison to sibs, profuse silk 
production in the tassel, a general lack of elongation in upper internodes, and early 
assumption of a pendant position by the tassel. In our stock, this pendant position 
is due to bending of the culm; in Tassel-seed 2, likewise a pendant form, it is 
attributable mostly to bending of the tassel proper. These differences are mani- 
festations of the degree of condensation (Anderson, 1944a), at least in part. 
Spikelet pairs of Tassel-seed 1 are generally closely spaced, resulting in rigid 
branches, while those of Tassel-seed 2 are more widely spaced, resulting in lax 
branches. Emerson mentioned these relative spikelet-pair densities, but he did not 
call attention to their differing effects on the area of bending. However, his figures 
10 and 11 show each form quite well. 

In our stocks, branch tips most commonly end in sterile zones characterized by 
closely overlapping glumes. Occasionally at the very tip, these glumes simulate 
those of male spikelets; they sometimes even contain stamens, but no pollen 
shedding has been observed. 


[Vol. 42 

The ear of Tassel-seed 1 is slow to develop, and, as in Emerson's strains, fails 
to set fruit unless the tassel is either removed or poorly pollinated. It is normal 
except that, just as in the tassel, both florets of each spikelet often develop, 
obscuring evidence of regular rowing. 

2. Tassel-seed 2 (ts 2 ) (pi. 23, fig. 2). 

Emerson (1920) originally described this mutant, located on Chromosome 1 
(Emerson et al y 1935), as "tassel-seed," and noted that the tassels were loose, like 
those of normal plants, with individual spikelets more or less separated. Our 
material never showed individual spikelets, but always paired spikelets. In Emer- 
son's specimens, spikelet pairs sometimes occurred more densely, "but not ear-like 
in any way." Although it rarely occurred, he noted that staminate flowers might 
develop with pistillate ones throughout the entire tassel, but he did not determine 
whether these male florets were functional. He also observed that glumes and 
"palae" (paleas and lemmas) of such male spikelets were long, narrow, and pointed, 
as in normal tassels, while in female spikelets these parts were shorter, broader, and 
more rounded. 

In our material, spikelets were either female or perfect, no male spikelets having 
been observed. On perfect spikelets, glumes were elongate and near normal. Also, 
each spikelet pair was associated with a cupule which was often elongated above 
spikelet pairs located on central spike tips. Kernels developed on nearly all spike- 
lets; in many, as with Tassel-seed 1, both florets were functional and two kernels 
were formed. 

Each tassel branch is thin, with spikelet pairs more widely spaced in its distal 
portion, but the base may be as thickly set as any found in Tassel-seed 1. Tassels 
of this mutant are pendant, but differ from those of Tassel-seed 1 plants in that 
the bending is mostly accomplished in the tassel rather than in the supporting 
culm. This point is discussed further with regard to Tassel-seed 1 above. 

Ears of Tassel-seed 2 develop to a degree depending on the fate of the tassel. 
If little or no successful pollination takes place, ears are developed rapidly, but 
always later than on normal sibs. If tassels are removed soon after their appear- 
ance, ears are formed about the same time as in normals. These results are in 
accord with the findings of Emerson. Like most other tassel-seeded forms, second 
florets in many spikelets set fruit, after the manner characteristic of Country 
Gentleman sweet corn (Weatherwax, 1916), so that regular rowing may be 
obscured. When ears develop, they are of normal size. 

3. Tassel-seed 3 (Ts 3 ) (pi. 24, fig. 1). 

The first published mention of this dominant mutant was made by Phipps 
(1928), who stated that "a third type of tassel seed designated as Tassel-seed 3 


of Emerson's. 

1935) also listed Tassel-seed 3 as being an unpublished 
They noted that this form was similar to Tassel-seeds 
inflorescence was generally mixed pistillate and staminate 

ually pollen could be obtained. The 



Tea Mays 201 

Tassel-seed 3 tassels are easily recognizable, but, like other forms, they are 
variable in appearance. Branches bearing distichously arranged pistillate spikelets 
resemble flat green ribbons on their adaxial surfaces; both branches and central 
spikes most often end in tapering sterile rudimentary pistillate spikelets. Often, 
these sterile spikelets have an appearance of being whorled rather than distichously 
arranged. About three-fourths of our specimens p 


oduced some spikelets from 
which the florets shed pollen. These staminate spikelets are imperfect, most com- 
monly produced on proximal portions of branches and central spikes and sometimes 
separated by sterile zones from pistillate regions. An occasional branch is entirely 
staminate and ends with a zig-zag axis, but most branches which bear staminate 
spikelets end in the characteristic pistillate tips described above. Tassels also are 

as they are exserted. The 
upper three to five internodes of Tassel-seed 3 plants do not elongate to the degree 
common in Standards, and as a result a rather tight collar of leaf sheaths is formed 
covering the lower sixth of each tassel. One tassel in our culture was infected 
with smut (Ustilago zeae) , a probable reflection of the extreme growth and lasting 
succulence of tassel parts. Emerson (1920) observed, and Dr. E. G. Anderson 
agreed, that Tassel-seeds 1 and 2 were highly susceptible to smut, but this year we 
noted no infections in any of our stocks except Tassel-seed 3. 

There were often instances in which a spikelet pair consisted of one staminate 
and one pistillate spikelet; in these mixed pairs, the staminate one is always pedi- 
cellate. No perfect florets were observed, and only one floret in any particular 
pistillate spikelet would form a caryopsis. Cupules are well- developed adaxial to 
both pistillate and mixed spikelet pairs. 

The ear of Tassel-seed 3 shows very few secondary florets, although Emerson 
et al (1935) mentioned that such florets do develop. In other respects, the ear 
resembles that of a Standard plant. 

4. Tassel-seed 4 (tsj (pi. 24, fig. 2). 

First reported by Phipps (1928), and also as Sorghum Tassel (Hayes and 
Brewbaker (1928), ts± was assigned by Emerson et al (1935) to Chromosome 3. 
Phipps described the tassel as being tassel-like in structure but predominantly 
pistillate-flowered, and stated that when it emerged it was a mass of silks. Our 
material showed tassels much shorter than normal, with an abnormal number of 
weak branches held upright by a tangled growth of silks. In Phipps' stock, mature 
tassels were studded with kernels, usually densely packed together, especially on 
the central axis of the tassel, and not arranged in regular rows. Our specimens 
varied from setting fruit to this extent to setting no fruit at all. Phipps interpreted 
this irregular rowing as a result of development of a second floret in each spikelet; 
with this finding we are in agreement. In both his material and in ours, tassel 
branches were short and had distichously arranged spikelets, the spikelets on the 
proximal branch and the lower central spike producing stamens which contained 

functional poll 

papery, but we cannot 


[Vol. 42 

Phipps that they were similar to glumes of a normal ear; we observed no glume 
induration. We noted also that sessile spikelets rarely branched, while pedicellate 
ones often formed short branches bearing two to six spikelets. Our standard stock 
has an average condensation index of 1.3; spikelets may occur at a given node in 
pairs, in 4*s, in 3's and 5's and occasionally in 6's. As in Phipps' stock, both florets 
of each spikelet always developed, and thus each node had at least four and oc- 
casionally as many as twelve florets in addition to possible extra ones formed on 
branches of proliferated pedicellate spikelets. Phipps mentioned that careful study 
of his material showed "a few multi-flowered spikelets." 

Silks were produced only from pistils, and such florets were always imperfect. 
No glumes were modified into silks, as Phipps reported. It is doubtful if his ob- 
servation is accurate, because he noted "the basal parts of such glumes were 
normal." The silk-like structures were most likely awns, and if so, they were 
probably produced not on glumes but upon lemmas. We observed no such struc- 
tures in this mutant, but in a mutant combination not reported on here, awns are 
developed on the lemmas. In some tassels, spikelets were sterile, producing neither 
silks nor stamens. These sterile spikelets were most common on branches between 
the proximally located obviously staminate areas and the distally located obviously 
pistillate areas. Characteristic of our Tassel-seed 4 stock were both its generally 
light set of fruit in the tassel and production of staminate and pistillate florets in 
widely varying amounts. No cupules were developed adaxial to any spikelet pairs 
of the tassel. 

Ears of Tassel-seed 4 were often shorter than normal because of failure of the 
tip to mature; they developed better if tassels were removed early. In our stocks, 
no excessive development of silks was noted, but Phipps reported that some glumes 
were modified into silks. Development of second florets, referred to by Phipps as 
common, was apparently much less common in our stock, and regular rowing was 
not obscured. Glumes were usually less indurated than in normal ears. 

5. Tassel-seed 4* (tsf). 

This allele of Tassel-seed 4 was found by E. G. Anderson in a background 
stock of CC5 X L317 which had been exposed to radiation at Eniwetok. Crosses 
made by the junior author between this mutant and Tassel-seed 4 showed the two 
types to be allelic. 

The tassel of Tassel-seed 4 a resembles that of Tassel-seed 4, but the quantity 
of seed set is much greater. Pollen is much more freely produced than in Tassel- 
seed 4. Pistillate florets which set fruit are always borne on pedicellate spikelets 
and are generally imperfect, but some perfect florets have been found which always 
appeared to set fruit. In general, pistillate florets are confined to the outer third 
of branches. The central spike is sometimes all staminate, with several extra 
florets produced on short branches which are proliferations of pedicellate spikelets. 
Cupules are sometimes weakly developed on the culm of the central spike, but 
their presence is not universal. Average lengths of glume are about half again as 
great as those of Tassel-seed 4. 



In Tassel-seed 4 a a few pedicellate spikelets have a tendency to proliferate 
slightly in the upper (distal) third of the ear. After kernels are removed, the right- 
angled insertion and excessive lengths of the glumes, paleas and lemmas of this 
allele give the cob a markedly different appearance from that of Tassel-seed 4. 

6. Tassel-seed 5 (Ts 5 ) (pi. 25, fig. 1), 

Although this mutant is attributed to Emerson (1932), the only mention of 
Tassel-seed 5 in his paper was its inclusion on Chromosome 4 of a linkage map 
(see his fig. 1, p. 145). Emerson et al (1935), in a brief description, noted that 
it contained both silks and anthers but was not compacted as in Tassel-seed 4. In 
our material, tassels were very close to normal proportions, but they were never 
exserted as far as in normal sibs. Apparently this condition was due to lack of 
elongation of the internode directly below each tassel, as the plants appeared to be 
normal in height (Table I and fig. 1). Silks are much shorter and more scattered 
than in previously discovered tassel-seed forms. Their occurrence varies from very 
few being located in small branch areas (either basal or terminal), to being uniform 
over the entire tassel. 

Spikelets may bear either staminate, pistillate, or perfect florets, but these 
types occur in specific places. Imperfect pistillate spikelets are usually located in 
proximal parts of lower branches, and when they occur on central spikes, it is only 
in their lower regions. Cupules are well developed adaxial to such spikelets, which 
most often occur as sessile members of spikelet pairs; the pedicellate ones are always 
both imperfect and staminate. The imperfect pistillate florets form most of the 
caryopses. Perfect spikelets are never associated with well-developed cupules, and 
seldom set fruit unless there are few or no imperfect pistillate spikelets present. 
Even here, however, fruits are formed predominantly on sessile spikelets. If a 
fruit is set in such a floret, the stamens are generally not exserted, but examination 
shows their anthers to be full of pollen and occasionally dehisced inside the palea 
and lemma. Imperfect staminate spikelets occur wherever the other two forms 
are absent; they never have any trace of silks, and their stamens shed pollen 
copiously. Silks not pollinated quickly withered. 

The ear of Tassel-seed 5 in our strain developed very few second florets, but 
Emerson et al noted that "secondary florets develop in ears." Otherwise, the ear 
closely resembles that of our Standard stock. 

7. Tassel-seed 6 (Ts 6 ) (pi. 25, fig. 2). 

The first apparent reference to this dominant mutant was made by Emerson 
(page 14 of the Maize Genetics Cooperation Newsletter for January, 1933). In 
subsequent 'Newsletters', he established that this mutant is located on Chromosome 
1 , but no description of its morphology has been published. 



Standards. In general 


[Vol. 42 

appearance, the tassels resemble those of Tassel-seed 4, but they usually set a 
considerably greater number of fruits. Branches are numerous, short, thin, and 
lax, with no cupules developed adaxial to any spikelet pairs. They end in many 
small florets which rarely form fruits and which may be borne on whorled instead 
of distichous spikelets. Spikelet pairs are closely set on both branches and central 
spike. Scarcely any distinction exists between pedicellate and sessile individuals, as 
each one forms several closely packed florets. Short branches with 4 to 12 spike- 
lets are developed on the central culm, giving it a thickened appearance. Glumes 
are hyaline, without chlorophyll, and extremely short, averaging about 5 mm. in 

Tassels of this mutant have two general forms. Approximately half of them are 
entirely pistillate; the others produce staminate spikelets on proximal branch areas 
and on the lower third of their central spikes. Tassels of this latter sort may have 
their remaining spikelets constructed in one of three different orders which occurred 
in our samples with equal frequency: (1) florets bearing silks might also bear 
stamens and thus be perfect; (2) they may bear no stamens and be imperfect, or 
(3) a zone of perfect florets of variable length may grade off into a tip which is 

>erfect. Stamens in perfect florets often neither extruded their anthers 
nor shed pollen, but those in staminate florets invariably shed good quantities of 
pollen. Silks are produced only from pistils, but an occasional lemma near a 
branch tip may develop a short awn. 

Ears of Tassel-seed 6 are not well developed unless the tassel is removed early. 
There is the same characteristic production of second florets found in other tassel- 
seeded forms, so that regular rowing is often obscured. In other respects, ears are 
comparable with those of Standard plants. 


8. Tassel-seed 7 (ts 7 ) (pi. 26, fig. 1), 

This mutant was found by E. G. Anderson at Pasadena in material which had 
been exposed to radiation at Bikini. (See Anderson et al, 1949, for a further dis- 
cussion of radiation effects.) It is not known to which chromosome it should be 


tions and dimensions of normal ones, and, like Tassel- 
ristillate florets. Silks are generally short and evenly 
varies from almost none to profuse. Each silk arises 
ret; florets which do not produce silks are imperfect 
and staminate, producing pollen freely. Only two florets are borne on each spike- 



the tassel. The number 

twelve per tassel, and on about half the tassels they are not formed at all. 

The ear of Tassel-seed 7 commonly bears up to six short branches at its base. 
These branches are produced from pistillate spikelets, and an adaxial cupule is 
present above each one. The sessile spikelet of such a pair is generally not fully 
developed. These basal branches do not form any viable kernels. Silks form 
only slightly if at all, even though the branches remain meristematic after cessation 
of growth elsewhere throughout the plant. Occasionally, two or three small 



branches are found on the shank below an ear. These sub-ear structures have 
always been sterile, and none of them developed beyond the point at which 
pistillate spikelets could be easily recognized. 

9. Tassel-seed 8 (ts 8 ) (pi. 26, fig. 2). 

This mutant, found by E. G. Anderson at Pasadena, was segregating in a cul- 
ture of chromosome translocations. It is known to be linked to white endosperm 
(y 1 ). Both ts 8 and ms 1 tassels are male-sterile, in chromosome 6 and linked to 
y 19 but the two forms are not allelic. Further, since Ts s sets no fruit but produces 
silks and since Emerson et al (1935) mentioned no male-sterile forms which so 
behaved, the material is here described as an eighth tassel-seed form. In over-all 
appearance, plants resemble normal sibs. Tassels are of standard proportions and 
size, but glumes are occasionally shorter than normal and the spikelets never swell 
as time for anthesis approaches. Silk production varies from none to a profusion 
as great as may occur in Tassel-seed 7. Four-fifths of our population of this 
mutant had tassels in which no stamens were formed; in the remaining one-fifth, 
they were formed but their anthers were empty. No stamens are ever exserted. 
Silks arise from abortive pistils. Three florets per spikelet may send out silks, but 
no fruits are ever set regardless of ample exposure to pollen. Branch tips are often 

The ear of Tassel-seed 8 produces a great profusion of silks; 4 to 6 are formed 
by each spikelet. The first silk formed (from the lowest flower) is of greatest 
diameter, and its pistil forms a caryopsis. Each of the others sends out a silk which 
generally is exserted, but no fruits are formed. Kernels are arranged in rows, and, 
except for the fact that many extra silks protrude between the tightly packed 
kernels, a mature ear resembles that of a Standard plant. 

10. Male-sterile (msj (pi. 27, fig. 1). 




tassels of male-sterile plants were more slender than normals, that anthers failed to 
extrude, and that no pollen was shed. They also reported that Beadle found 
meiosis essentially normal but that the haploid nucleus did not go through the first 
mitotic division, and by the time pollen should be shed it was difficult to find even 

of spores. Emerson et al noted further that shriveled anthers were 
exserted much later than in normal sibs. 

In our material, no stamens were exserted. In other respects, tassels of ms 1 
resembled those of our Standards. There was a tendency for the tassel to 
emerge sooner than with normals, and this tendency was reflected in slightly longer 




[Vol. 42 


The ear is normal. Singleton and Jones reported that ms 1 is closely linked to 
the white endosperm locus ( y x ) with about 5 per cent recombination. Our material 
behaved in the same fashion; white kernels almost invariably produced male-sterile 

11. Brancbed-silkless (bd) (pi. 28, fig. 2.) 

First described by Kemp ton (1934), this mutant is listed by Emerson et al 
(1935) as occurring on Chromosome 7. Kempton noted that the division of the 
tassels into a central spike and branches is as definite as in normals, but that the 
tassels have a thickened appearance suggesting those of Tunicate plants. He at- 
tributed this thickening to development of short branches in place of paired 
spikelets, a condition which also occurred in our material. He also mentioned the 
fact that this branching was most common on central spikes, and that on the 
branches there was more tendency toward retaining a pedicellate-sessile form. 

Kempton made no mention of sterility, nor did his illustrations show any, but 
in our material sterile areas occurred at branch tips and sometimes at the tip of the 
central spike. Not uncommonly lowermost branches were sterile throughout. In 
sterile areas, spikelets were less developed as distance from the central culm in- 
creased. Kempton noted that many plants shed pollen in good quantity, but 
stamens often were not fully exserted. 

Ears of Branched-silkless are of two forms, a fact which Kempton likewise 
noted; his figures show each type clearly. In one type, each spikelet primordium 
develops into a short branch bearing female spikelet rudiments which never develop 
beyond the earliest stages. In the other type, such branch development is con- 
fined to the basal quarter of the ear, the upper three-fourths being composed of 
sterile spikelets with elongated and sometimes thickened glumes. That these 
branches develop from spikelet primordia on an otherwise normal cob can be easily 
shown by removal of the branches; underlying each pair adaxially is a well- 
developed cupule. It is this latter type which may occasionally produce a scatter- 
ing of silks too short to become exserted. Kempton stated that his material was 
wholly without silks, but Emerson et al mentioned that the ear occurs "often 

without silks. 


12. Club (club) (pi. 27, fig. 2). 

This mutant was first noted by E. G. Anderson and segregated from non- 
irradiated genetic stocks at the Caltech farm in Arcadia. Its chromosome length 
is unknown. 

The tassel is shorter than normal, with branches held at angles approximating 
45°. Branch ends do not droop and their thickly set spikelet pairs are fertile 
throughout. The central spike is exceedingly thick, and it is on account of this 
character that the mutant was named. This thickening is brought about by a 
shortening of internodes, so that the spikelets stand nearly at right angles to the 
central culm. Further, these closely spaced spikelets occur singly or in pairs; there 
is no pedicellate branch development such as is found responsible for creation of 
thickened portions of the central spike of Branched-silkless. 



Tea Mays 207 

The ear of Club generally matures a week later than the tassel. Its silks are 
profuse, and there are 4-6 silks per spikelet, each from a rudimentary pistil — a 
condition also found in ears of Tassel-seed 8. However, Club ears are fascia ted, 
the branching strongly resembling a type listed by Kemp ton (1923) as Bearsfoot 
(see his fig. 1 for a clear illustration) . It is apparently caused by one or more 
incomplete divisions of the growing point when the ear is partly formed, followed 
by simultaneous development of each new point into a more or less independent 
ear tip. The upper half of such an ear is generally hollow, and no spikelets are 
formed on the inside walls. 

13. Vestigial glume (Vg) (pi. 28, fig. 1). 

First reported by Sprague (1939) , Vestigial glume was found to be a dominant 
mutation on Chromosome 1. It is easily noted in the tassel, where the hyaline 
outer glumes range from awl-shaped vestiges to nearly one-third normal length. 
Sprague pointed out that "flowering glumes" (lemmas) and "palets" (paleas) 
were also reduced in size, with the result that the stamens were nearly completely 
exposed. In our material, as in his, anthers generally dried up before dehiscence, 
but occasionally a plant shed viable pollen. Many specimens in our stock had 
sterile areas (i.e., no stamens were formed) at tips of branches and central spikes, 
in some cases involving the outer sixth of all branches. Generally, over-all tassel 
size was smaller and tassel construction, as exemplified by thickness of central 
culm and branches, was lighter than in normal sibs. 

The ear of Vg is easily recognized, as Sprague pointed out, after removal of 
kernels. There are no chaffy upper and lower glumes, paleas or lemmas present, 
only short pedestals, each with a low, hard ridge on both upper and lower sides. 
Above each spikelet pair, in an adaxial position, is a cupule with somewhat reduced 
rachis-flaps (Lenz, 1948). 

14. Tunicate (Tu) (pi. 22, fig. 2). 

According to Collins (1917a), this mutant has been reported from several 
sources and its origin is not known. Recently, it has been cited in connection with 
theories of maize origin (Mangelsdorf, 1948; Mangelsdorf and Smith, 1949). 
Cutler (1944) suggested that its widespread occurrence among Indian tribes was 
because of mystical significance attached to it. Emerson et al (1935) listed the 
character as occurring on Chromosome 4. 

Tassels of heterozygous Tunicate plants have normal measurements in all 
visible parts except the glumes. The glumes are one-fifth to one-fourth longer 
than those on Standard tassels and enclose paleas and lemmas similarly elongated. 
The stamen length being normal, the anthers, especially in the central spike, are 
not fully exserted and therefore are not pendant at anthesis. Pollen is shed 


Ears of heterozygous Tunicate plants are easily classified. Collins (1917a, b) 
stated that "the glumes of the female inflorescence, or ear, are developed so that 
each seed [fruit] is entirely enclosed." Emerson et al (1935) likewise attributed 

[Vol. 42 


these covers to glumes. In our stock, paleas and lemmas were also elongated, and 
in many basal spikelets, paleas and lemmas of both upper and lower florets were 
noted. No fruit occurred in these lower florets, however, and regular rowing was 
externally apparent. The rachis of a Tunicate cob is rather flexible. Cupule 
development adaxial to each spikelet pair was not excessive in comparison with 
Standards; but in the Tunicate ear of Guarani maize sent to the senior author by 
Dr. P. C. Mangelsdorf, the cupules are much more apparent than is normally the 


Homozygous Tunicate plants, according to Emerson et al (1935), are usually 
female-sterile. No homozygous plants were available in our collection. 

An allele of Tunicate, tu h , was reported by Mangelsdorf (1948) as being 
present in Maiz chapolote. How this allele would behave with our Standards as 
background was not determined. 

15. Ramosa 1 (raj (pi. 29, fig. 1). 

This form was originally described by Gernert (1912) under the name of Zea 
ramosa, and was listed by Emerson et al (1935) as Ramosa-ear 1, located on 
Chromosome 7. Gernert's description of the tassel was brief. In his stocks, tassels 
were slightly smaller than normal, invariably much branched and cone-shaped. 
Collins (1917a, b) noted that branches were much more numerous than normal. 
Kemp ton (1921) reported 400 as an extreme number; the branches gradually de- 
creased in size upward, the transition from branches to pairs of spikelets being 
imperceptible. A short central spike was characteristic in our specimens — a fact 
which Kempton had likewise noted in his plants but which Collins did not mention. 
It was not possible, according to Collins, to distinguish between plants heterozygous 
for Ramosa 1 and normals. However, in our material, it was quite simple to sep- 
arate them; in five F x families of Ramosa 1 X Standard, four with 10 plants and 
one with 9 plants, separation was easily made into 24 normal and 25 heterozygous 
individuals. Resemblance of tassels of homozygous Ramosa 1 plants to grass 
panicles is striking (Kempton's plate 13 is typical). Tassels of plants heterozygous 
for this character were intermediate between a pyramidal paniculate tassel and a 
normal one, identical to one figured by Kempton in his plate 14. 

The ear of Ramosa 1 was described by Gernert as being much branched, with- 
out male florets, covered with husks, and composed of a mass of kernels borne on 
numerous irregular branches. His description applies to our specimens as well, 
except for two other points: (1) branches on a Ramosa 1 ear are about as num- 
erous and have much the same irregular whorled arrangement as do those of the 
tassel; (2) there are no adaxial cupules at the junctions of ear branches and 
central culm. Cupules were present, however, adaxial to each spikelet pair of the 
branches. No other ear among the mutants in this collection is so organized. It 
was also a common occurrence for tips of those branches originating near the upper 
part of the ear to be sterile. 



7*ea Mays 209 

16. Ramos a 2 (ra 2 ) (pi. 29, fig. 2). 

The only published reference to this mutant is found in Emerson et al (1935), 

which simply listed Ramosa 2 as being located on Chromosome 3 and credited its 
discovery to Brink. 

Ramosa 2 has a tassel characterized by stiff upright branches which remain 
closely appressed to the central culm. The central spike region does not bear short- 
pedicelled spikelet pairs; instead the spikes are borne on branches which decrease 
in length from base to apex. There is a pronounced transition from well-developed 
tassel branches bearing many pairs of spikelets to smaller branches bearing fewer 
spikelets, some of which appear to occur singly. As one proceeds acropetally, 
these multi-spikeleted branches are replaced by stalks each bearing only one 
pair of spikelets which are themselves borne on pedicels longer than normal. In 
the adaxial area of each of these branches and stalks, a small cupule-like depression 
in the central culm can be observed which is generally of a different color from 
the surrounding surface. Although these depressions and short stalks are not 
confined to Ramosa 2, they are here most common and best developed. The spike- 
let pairs in proximal areas of well- developed lower branches are likewise stalked 
and also have abnormally long pedicels, but there is no adaxial depression. 

Pollen is shed freely, but much of it remains within the confines of the tassel 
because close proximity of branches does not allow unrestricted air passage. Each 
spikelet has two male florets, and each of these has three functional stamens. 
Tassel-seed 4 a might possibly be confused with Ramosa 2, as its tassels sometimes 
show the same tendency to have numerous stiff, erect branches and stalked spikelet 
pairs in the central culm region, with an abundance of pollen produced. It can 
be separated from Ramosa 2 by the lack of stalked spikelet pairs on lower branches 
and the fact that each of its branches end in a few sterile undeveloped spikelets. 

The ear of Ramosa 2 sometimes ends in a staminate structure looking like a 
normal central spike. Most commonly, it produces scattered branches on the 
upper (distal) half of the cob, which in turn bear female spikelets in pairs. These 
branches are nearly always found in younger ontogenetic stages than the rest of 
the plant and consequently set seed only rarely. A branch originates by growth 
of what would normally be the pedicellate spikelet of a pair of kernel-bearing 
spikelets. No instances were observed in which the sessile spikelet of such a pair 
formed anything but a normal caryopsis. The spikelet pairs themselves are borne 
on short stalks; the cupule is bent at nearly right angles, and is adherent half to 
the stalk and half to the cob. In addition, these stalks are spaced rather widely 
apart, so that a cob of Ramosa 2 from which the chaff has been removed resembles 
a similarly treated cob of Coroico maize (Cutler, 1946). 


Morphological studies have been made on tassels and ears of sixteen mutant 
forms of maize. Four of these forms are reported for the first time, and the 
others, for which descriptions were either inadequate or non-existent, are here 

[Vol. 42 


described. Each genetic form was introduced into a standard background derived 
from the inbreds CC5 and L317. The mutants studied included nine tassel-seeded 
forms: ts 19 ts 2 , Ts 3 , ts 4 , /s 4 a , Ts 5 , Ts 6 , ts 7 and ts 8 . Others studied were Male- 
sterile (msj, Branched-silkless (bd), Club (club), Vestigial glume (Vg) f Tuni- 
cate (Tu), Ramosa-1 (raj and Ramosa-2 (ra 2 ). 


Alava, Reino O. (1952). Spikelet variation in Zea Mays L. Ann. Mo. Bot. Gard. 39:65-96. 
Anderson, Edgar (1944a). Homologies of the ear and tassel in Zea Mays. Ann. Mo. Bot. Gard. 

, (1944b). The sources of effective germ -plasm in hybrid maize. Ibid. 3 55-3 61. 

, (1949). Introgressive hybridization. 109 pp. New York. 

, (1951) The sacred plume. 24 pp. Publ. Pioneer Hi-Bred Corn Co. Des Moines. 

, and W. L. Brown (1952a). The history of the common maize varieties of the United 
States corn belt. Agric. History 26:2-8. 

» ■ (1952b). Origin of corn belt maize and its genetic significance. In: Heterosis, 

pp. 124-148. Edited by John W. Gowen. State College, la. 

Anderson E. G., A E. Longley, C. H. Li, and K. L. Retherford (1949). Hereditary effects pro- 
duced in maize by radiations from the Bikini atomic bomb. I. Studies on seedlings and pollen 
of exposed generation. Genetics 34:639-646. 

Collins, G. N. (1917a). Hybrids of Zea ramosa and Zea tunicata. Jour. Agric. Res. 9:3 83-395. 

- , (1917b). Hybrids of Zea tunicata and Zea ramosa. National Acad. Sci. Proc. 3:345-349. 

Cutler, H. C. (1944). Medicine men and the preservation of a relict gene in maize. Tour. Hered. 
3 5:291-294. J 

, (1946). Races of maize in South America. Harv. Univ. Bot. Mus. Leafl. 12-2 57-292 
Emerson, R. A. (1920). Heritable characters of maize. II. Pistillate-flowered maize plants. Tour. 
Hered. 11:65-76. r J 

, (1932). The present status of maize genetics. Sixth Internat. Cong. Genetics Proc. 



"'iP;r W * ? Cadle ' 3nd A ' C Fraser < 1935 )- A summary of linkage studies in maize, 
nell Univ. Agric. Exp. Sta. Mem. 180:1-83. 

Gernert, W. B. (1912). A new subspecies of Zea Mays. Amer. Nat. 46:616-622. 
Hayes, H. K., and H. E. Brewbaker (1928). Heritable characters in maize. XXXIII. Sorehum 
tassel. Jour. Hered. 19:561-567. 

Kempton, J. H. (1921). Inheritance of ramose inflorescence in maize. U. S Dept Aer Bull 
971:1-20. 5 " 

-, (1923). Heritable characters of maize. XIV. Branched ears. Jour. Hered. 14:243-252. 

, (1934). Heritable characters in maize. XLVII. Branched silkless. Ibid. 25:29-32. 

Lenz, L. W. (1948). Comparative histology of the female inflorescence of Zea Mays L Ann Mo 
Bot. Gard. 35:353-376. 

Mangelsdorf, P. C. (1948). The role of pod corn in the origin and evolution of maize. Ann. Mo. 
Bot. Gard. 3 5:377-406. 

, and C. E. Smith, Jr. (1949). New archaeological evidence on evolution in maize Harv. 

Univ., Bot. Mus. Leafl. 13:213-247. 

Nickerson, N. H. (1953). Variation in cob morphology among certain archaeological and ethno- 
logical races of maize. Ann. Mo. Bot. Gard. 40:79-111. 

— - , (1954). Morphological analysis of the maize ear. Am. Jour. Bot. 41:87-92. 

Phipps, I. F. (1928). Heritable characters in maize. XXXI. Tassel-seed 4. Jour. Hered. 19:399-404. 

Singleton, W. R and D. F. Jones (1930). Heritable characters in maize. XXXV. Male sterile. 
Jour. Hered. 21:266-268. 

Sprague, G. F. (1939). Heritable characters in maize. L. Vestigal glume. Jour. Hered. 30:143-145. 
Weatherwax, Paul (1916). Morphology of the flower of Zea Mays. Bull. Torrey Bot. Club 43:127- 



Fig. 2. Drawings to same scale of spikelet pairs found on lowermost primary tassel branches 
of certain mutants showing characteristic features: it% 9 adaxial view of thick, ribbon-like branch 
showing cupules and a developed second floret (silks removed). Ts? y , pedicellate (right) and sessile 
(below) spikelets with hyaline awnless glumes; pedicellate spikelet forms numerous naked small 
pistils, bd, both spikelet axes form extra spikelets. r*2, an extra interncde is found between the 
primary axis and point of departure of the sessile spikelet. Vg, glumes do not develop, so stamens 
are left naked. ts 4i sessile spikelet left has one male and two female florets; pedicellate spikelet 
(right) forms numerous spikelets, with many of the thin glumes ending in soft awn-like tips. 
rs 3 , sessile female and pedicellate male in one spikelet pair; pedicellate axis adheres to edge of cupule 
for part of its length (silks removed). 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 22 






















Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 23 















2 1 












Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 24 


h ■ 





* X 



* V J£ 


































Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 25 
























Ann. Mo. Bot. Gard., Vol. 42, 195 5 

Plate 26 

































* i— • 



Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 27 







r 1 









o . 

2 crc* 







...- . 



Ann. Mo. Bot. Gard., Vol. 42, 195 5 

Plate 28 





























r- ^ 

E * 

2 Z 


Ann. Mo. Bot. Gard., Vol. 42, 195 5 

Plate 29 













2 1 




Z £ 









In the course of examination of specimens in the fern collection of the 
Botanisch Museum en Herbarium, Utrecht, the senior author discovered a speci- 
men of Doryopteris that seemed to be different from any species treated in the 
junior author's revision of the genus 1 . Further examination has proved this to 
be the case and it is here described as new. 

Doryopteris conformis, spec. nov. 

Rhizoma modice crassum breviter repens, squamis elongatis angustissimis dense 
vestitum, partium hyalinarum cellulis latitudine maxime partem quintum longi- 
tudinis aequantibus; stipes obscurus laevis vel leviter rugosiusculus, fasciculis 
vascularibus duobus; lamina fertilis sterili similis coriacea, venatio libera; sterilis 
suborbicularis — quinquangularis, profunde bipinnatifida; fertilis conformis, pro- 
funde bi- vel tripinnatifida; receptaculum intramarginale plus minusve continuum; 

sporangia breviter pedicellata. 

Typus: Gonggrijp & Stahel (B. W.) 5699; Surinam: Mt. Hendriktop, alt. 1080 

m.; moist, sunny rocks; in Herb. Utrecht. 

Rhizome moderately stout, short-creeping; scales of the apex of the rhizome 
very long and narrow, the cells of the hyaline portions at least five times as long 
as broad; stipe dark purple to black, naked or slightly scaly at the base, glabrous, 
smooth or minutely and irregularly roughened, with two vascular bundles at the 
base, terete; fertile and sterile blades similar, without proliferous buds, coriaceous; 
sterile leaf about 25 cm. long; blade 7 cm. long, suborbicular-pentagonal, deeply 
bipinnatifid with about 12 oblong, broadly rounded, entire or partially crenulate 
ultimate lobes; margin with a pale brown cartilaginous border; venation free; 
hydathodes prominent on the upper surface; fertile leaf about 25 cm. long; blade 
about 8 cm. long, suborbicular-pentagonal, deeply bi- to tripinnatifid, with num- 
erous ultimate segments; primary segments broadly decurrent and surcurrent, the 
bases forming wings along the rachis with symmetrically concave sides; ultimate 
segments oblong-lanceolate, narrowly rounded, entire; soral lines continuous around 
the sinuses; sporangia short-stalked, i.e. the stalk somewhat shorter than capsule 
(which is 340—360 fx long), borne on a more or less continuous vascular commis- 
sure; spores subglobose, triplanate, slightly rugose, pale yellow-brown, about 55 fi. 

Doryopteris conformis is most closely related to D. lomariacea from which it 
differs in the non-dimorphic fertile and sterile leaves, the segments of the fertile 
being much broader than in D. lomariacea, in the short-stalked sporangia and 
coriaceous texture of the blades; in D. lomariacea, the sporangia are long-stalked 

and the leaf-tissue is herbaceous. 

This species is notable in that it is the only local endemic in the genus outside 
of southeastern Brazil. However, the Guiana Highlands represent a weak sec- 

1 Tryon, Rolla M. A revision of the genus Doryopteris. Contr. Gray Herb. 143:1-80. 1942. 
♦Issued November 17, 1955. * ' 

[Vol. 42, 1955J 


ondary center for the genus in South America. This is the sixth species known 
from there, not including the widespread and somewhat doubtfully allied D. 
concolor; a concentration of species exceeded only in the southeastern Brazilian 
Highlands. It is also significant that all the other five species, D. lomariacea, 
sa git ti folia, collina, v avians, and pedata var. multipartita are variously disjunct 
between southeastern Brazil and the Guiana Highlands. Whether D. conformis 
represents a local offshoot of D. lomariacea or whether it was evolved, as all other 
Guiana species evidently were, in southeastern Brazil and migrated via the Andes 
to the Guianas, can onlv be determined if it is eventuallv discovered in Uracil. 

Explanation of Plate 


Fertile and sterile leaf of Doryopteris conformis, from type. X % 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 30 





There are about 500 species which have been referred in the past to the genus 
Celastrus; of these half are African plants. In 1942 Loesener 1 , in a review of the 
genus for Engler & PrantPs Tflanzenfamilien', transferred some of the African 
species to the genus Gymnosporia and some to the allied genus Maytenus. In 
indicating the geographical distribution of Celastrus only in Asia, America, 
Australia, and Madagascar, he seems to have been aware of the fact that the so- 
called species of Celastrus of Africa proper did not truly belong to that genus. 

Loesener has expressed the opinion that the confusion regarding species limits 
in Celastrus and the synonymy indicate the need for a more precise study. I have 
attempted in this treatment to define the generic limits of Celastrus and its rela- 
tionship with other closely related genera, to review and to check all the published 
binomials of Celastrus, and to clarify the complicated synonymy. I have also made 
a study of the morphological characters and geographical distributions of the dif- 
ferent species. I have used the data to separate species or groups of species and to 
show their inter-relationships (fig. 1). All the (f Celastrus y> species from Africa 
proper have been excluded in this treatment, as they are referable either to Gymno- 
sporia or Maytenus. I have tried to delimit definitely the three genera, Celastrus, 
Gymnosporia, and Maytenus, which have been very much confused. It is my 
intention in the near future to make a detailed study of these last two genera. This 
treatment of Celastrus includes two subgenera with thirty-one species and five 


The generic name comes originally from Theophrastos, who, however, desig- 
nated with the Greek word KeXaorpos (Kelastros) an evergreen tree (Phillyrea), 
that has nothing to do with our genus 2 . It is feminine in gender as used by Theo- 
phrastos, but Linnaeus in adopting it, made it masculine. It has already been 
pointed out by Airy Shaw 3 that under the International Rules the masculine gender 
must be retained. Hence, in this treatment, Celastrus is treated as masculine and 
I have standardized all the epithets accordingly. 

1 Loesener, Th., in Engler, A. & PrantI, K. Die Natiirlichen Pflanzenfamilien 2 Aufl. 20b: 131, 

134. 1942. 

2 Loes. loc. cit. 2 Aufl. 20B:132. 1942. 

3 Airy Shaw, H. K. in Curtis's Bot. Mag. 158:/. 9394. 1935. 

♦An investigation carried out in the Henry Shaw School of Botany of Washington University 
and submitted as a thesis in partial fulfillment for the requirements for the degree of Doctor of 


** Arnold Arboretum of Harvard University, Cambridge, Mass. 

Issued November 17, 1955. 


I Vol. 42 



The genus Celastrus was founded and described by Linnaeus 4 in his ^Genera 
Plantarum* (1737). Later, in his 'Species Plantarum' 5 (1753), he described three 
American and two South African species. Only one of these, Celastrus scandens, 
is now retained in the genus Celastrus. 

In 1824, Kunth 6 pointed out that some of the Celastrus species, for example, 
Celastrus buxifolia L., C. montana Roxb., C. trigynus Lam., etc., have peculiar 
characters and might well constitute a new genus. He gave a detailed description 
for that group of species, but provided no name for it. In 1834, Wight and 
Arnott 7 studied the Indian species of Celastrus, and, chiefly on the basis of ovule or 
seed characters, divided the genus into two sections, eucelastrus and gymno- 
sporia. They gave each section a very concise and clear description. In the sec- 
tion eucelastrus, the ovary is free from the disc, the ovules have a cup-shaped 
aril at their base, and the seeds are surrounded by an entire fleshy aril. In the section 
gymnosporia, which corresponds to Kunth's unnamed genus, the ovary is half- 
immersed in the disc, the ovules are naked at the base, and the seeds are apparently 
without an aril or with a very short, imperfect, and membranaceous one at the 
hilum. In 1862, Bentham and Hooker 8 elevated the section gymnosporia to the 
rank of genus. At present the limits of the genus Celastrus are identical with 
those of section eucelastrus of Wight & Arnott. 

Generic Relationships 


They often have been confused with Celastrus both in the literature and in the 



Maytenus, on the presence in the former of either thorns or inflorescences borne on 
short shoots, appears artificial. He says: "There seems little point in keeping the 
two genera distinct." 

After examining all available specimens of Celastrus, Gymnosporia, and May- 
tenus, I have concluded that they are distinct though very closely related genera. 

The following table for distinguishing the three genera is based on a consideration 
of several characters taken together and not any one single character taken by 
itself (pi. 31). 

4 Linnaeus, C. Genera Plantarum. ed. 1. 59. 1737. 

5 - ■• Species Plantarum. ed. 1. 195-197. 1753. 

6 Kunth, C. S. in HBK. Nov. Gen. & Sp. Pi. 7:64. 1824. 
7 Wight, R. & Arnott, G. A. W. Prodr. Fl. Ind. Orient. 152. 1834. 
8 Bentham, G. & Hooker, J. D. Genera Plantarum 1:364. 1862-67. 
9 Exell, A. W., in Kew Bull. 1953:103. 1953. 
10 Loes. loc. cit. 2 Aufl. 20b:109. 1942. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 31 










































































Celastrus Maytenus, and Gymnos porta contrasted. 


[ X. 42 


habit: — Plants of Celastrus are scandent, from 1 to 50 meters in height. Th 
branches are hardy, rapid-growing and vigorous. There are very few available 
records of the diameter of older branches for all the species, but two records of 


The branches and 

branchlets usually are terete, except in Celastrus angulatus, and are glabrous except 
on young branchlets, for example, of Celastrus hirsutus. 

Most older stems have orbicular to ovate lenticels. In Celastrus hindsii, lenticels 
are usually absent on the current year's growth. They are fine and dense in 
Celastrus angulatus and C. lenticellatus 9 but in C. hirsutus and C. c ase ariij olius they 
are large, elevated, ovate, or orbicular, and are found on the peduncles and pedicels. 
This feature may well serve as a supporting character to assist in identification of 
some species. 

vegetative buds: — Celastrus gemmatus bears axillary buds which are conical 
and characteristically large. All other species bear small, depressed and ovoid ones. 
The inflorescences of the species of series axillares (except Celastrus monospermus y 
C. monospermoidesy and C. hindsii) and one species of the subgenus racemo- 
celastrus, C. panamensis 9 are associated with accompanying vegetative buds, 
while the inflorescences of other species are not associated with vegetative buds. 
The outermost bud scales are usually deciduous, occasionally persistent, falcate, 
and spiny. These persistent bud scales are called "stipules" by some authors. 
Usually in Celastrus aculeatus and in some specimens of other species, there are 
four or more outermost bud scales as opposed to the usual two. These are nearly 
triangular, spiny, and semi-persistent or persistent. 

leaves: — The leaves are extremely variable, in the same species, and even on 
the same plant, the result of environmental conditions and the age of the plant. 
Previous authors mainly have used leaf characters for distinguishing species and 
varieties in the genus, but such characters are very unreliable. Most of the species 
are deciduous, but there are evergreen ones, for example, Celastrus monospermtis, C. 
monospermoides, and C. hindsii. The leaf shapes vary from elliptic to oblong, or 
from broadly ovate to orbicular. The apex and the base vary from acute to obtuse 
or rotund. The margins usually are serrate, serrulate or subentire; in Celastrus 
flagellaris they appear finely cilia te. In size, the leaves range from 2.0 cm. long 
and 0.8 cm. wide in Celastrus punctatus to 16 cm. long and 16 cm. wide in C. 


Texture usually varies from delicately to firmly membranous. In general, the 
texture of the leaves in flowering and fruiting stages of the plant is different: for 
example, in Celastrus angulatus the leaves of flowering specimens are delicately 
membranous, while those on the fruiting specimens are firmly membranous (so- 
called "coriaceous" of some authors). 

The leaves are glabrous or pubescent on the veins below. Sometimes they are 
pubescent in the juvenile stage and later become glabrous. In Celastrus hirsutus 
there is a dense brownish pubescence on both surfaces. 




The leaves of all the species have pinnate and netted veins. The midrib usually 
is elevated below, and distinctly or slightly elevated above- The primary lateral 
veins are four to nine pairs on each side of the midrib, arcuate towards the apex. 
The veinlets though usually visible, are obscure on both surfaces in Celastrus 
aculeatus, and while usually loosely reticulated are densely reticulated in C. hindsii 
and C. gemmatus. 

The phyllotaxy always is alternate. The petioles are canaliculate, from 0.5 to 
3.0 cm. long. In Celastrus flagellaris the petiole is half as long as the blade while 
in the other species it is less than half as long. 

floriferous branches and inflorescences: — Diagrams of the floriferous 
branches of the current season's or year's growth of most of the species of Celastrus 
have been drawn from herbarium specimens (pis. 32 and 33). These diagrams 
are arranged according to the sequence of the species in the key, and are not in- 
tended to show any evolutionary paths. They will make it convenient for a com- 
parison of the relationships among the species. They also reveal that the male and 
female plants of the same species have different inflorescence patterns. 

A "flower-cluster" has been thought of as a whole inflorescence or as a part of 
the inflorescence. Parkin 11 has commented: "It seems futile to quibble over the 
question whether the inflorescence means the mode of floral branching or the 
flower-group itself. Custom has sanctioned the latter meaning." I have adopted 
the customary approach of considering a "flower-group" as an inflorescence only 
as a matter of convenience in this treatment of the genus Celastrus. 

Dichasium and Aggregate Dichasium: — The fundamental type of inflorescence 
in the genus Celastrus is an axillary, solitary dichasium. A dichasium in its simplest 
form consists of three flowers — one terminal and two lateral, the terminal flower 
blooming first — and is organized as follows: First, there is the primary peduncle 
terminating in three secondary peduncles to each of which the pedicels of the flow- 
ers are articulated. What appears at first to be the entire flower stalk therefore is 
really a part of the peduncle and the whole pedicel. The dichasium usually is com- 
pound, and sometimes the branches are much multiplied to form a terminal panic- 
uliform aggregate dichasium. The aggregate dichasium differs from the thyrse in 
all axes being determinate and lacking any association with vegetative buds. In the 
literature the paniculiform floriferous branches are described usually as "panicles." 

Each dichasium may or may not be subtended by a foliage leaf or bract, but in 
the series axillares it is always associated with a dormant bud in the axil of the 
subtending appendage (pis. 31 and 32). All the species of subgenus racemo- 
celastrus have axillary dichasia with or without distinct primary peduncles. The 
Latin American species of racemocelastrus, in addition, usually have dichasia 
arranged to form a racemiform floriferous branch on the current year's growth 
(pi. 33). In this case the racemiform branch may be compounded and be located 
terminally on the branchlets to form a pseudopaniculiform aggregate dichasium 


Parkin, J. The evolution of the inflorescence. Jour. Linn. Soc. Bot. 42:512—563. 1914. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 32 

ftali.4* 79 if) EMl V (Fr. J 


Cl«Mn« bl&2 <T«*n« ft.) 





VLiM •* 

tf\ am»T t&M fttlltut 'fr.> 




llliM 36* f<?) Chov i»T <r>.) 





*%%*% aeaoo 

Ul "0 |f> 


VlUto Pr.J 

MtftMMi, f» £) r»m«t »40t *#) s.r.iu aoow »rr.> 




rr. » ■urn* iat» i£i ■ii«o« *♦« irr.) 


Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 33 

Porrast 16M1 CMftg 20682 <-£) 




Cull, to M.B.O. 

5*»ortso* f^) Vllton 7«12 f 






r eta lot 80M fcj) Till 60499 tFr.l 

Hanry MM (fl tl 1181 IPr.) a 



'tiara SaSO (<?) Poo 21«M tf-) 




rurpua T04t 


mm*** toto 


Sato-tear »• ■ 


- axillary bud 

- floral bud 


9 - blooming flower 



- fruit 

- bract 

- leaf 

Plates 32 and 33. Diagrams of the floriferous branches of current season's or year's growth of 
C el as t rus species. 

[Vol. 42 


(pi. 33). This type of dichasium differs from the paniculiform aggregate dichasium 
in series paniculati in having a vegetative bud in the axil of each dichasium. 

flowers: — The flowers of this genus are small and inconspicuous, usually about 
3-5 mm. long. According to the available records the flowers are yellowish-green 
or white. They are generally dioecious in the subgenus celastrus, and usually 
bisexual in the subgenus racemocelastrus. 

Calyx: — The calyx is campanulate and persistent, and sometimes accrescent as 
in Celastrus aculeatus. It always has five short and equal, deltoid, ovate, or oblong 
lobes which are usually imbricate or in some species rarely open in aestivation. The 
lobes have simple venation and are glandular- cilia te, slightly erose, or entire; they 
are glabrous or slightly pubescent on the outer surface, and reflexed or spreading 

in the fruit. 

Petals: — The flowers have five petals which usually are oblong or elliptic, 
glandular-ciliate to slightly erose, or entire. In Celastrus pantculatus the petals 
have parallel veins while in other species each of the petals has a "costa" and some 
lateral veins. They are generally glabrous, but in Celastrtis kusanoi are slightly 
puberulent on the inner surface. The petals usually are reflexed. 

floral Tube and Disc: — In the flower, there is a cup-shaped or plate-like struc- 
ture which is usually referred to as the disc. Berkeley 12 , who made a morphological 
and anatomical study of the flower of two species of Celastrus, C. scandens and 
C. orbiculatus, on the basis of the behavior of the vascular supply to the flower, 
interpreted the cup-shaped structure as the floral tube consisting of fused stamen, 
petal, and sepal bases. He considered the inner layer of the floral tube to be com- 
posed of fused basal portions of existing stamens and of stamens no longer present; 
between the existing stamens, there are small projections of tissue of reduced 
stamens constituting the "disc" in his sense. 

The cup-shaped and membranous "floral tube" is common in most of the 
species in this genus. It surrounds the pistil but is free from it; on the rim of the 
tube are attached the free floral parts. A plate-shaped, flat and fleshy "floral tube" 
is usually present, chiefly in all the species of subgenus racemocelastrus and 
several other species of the subgenus celastrus. The ovary is situated at the cen- 
ter of this type of "tube" and free from it, while all the floral parts are attached 
on the outer edge of the tube. The two species that Berkeley had studied lacked 
the flat floral tube, and it is difficult to interpret its origin without morphological 
and anatomical evidence. 

The term disc connotes the "floral tube" of Berkeley, and disc-lobe is equivalent 
to the "disc" in Berkeley's sense in the present treatment of Celastrus; this is just 
for convenience. The disc-lobe may be distinct or obscure, deltoid, oblong or 
reniform. Sometimes these are used as supporting characters to identify species. 

12 Berkeley, E. Morphological studies in the Celastraceae. Jour. Elisha Mitchell Scientif. Soc. 
70:185-206. 1954. 



Fertile and Sterile Stamens: — There are five stamens in the staminate flowers. 
The stamens depart from the rim of the cup-shaped disc, or are attached immedi- 
ately beneath the edge of the flat disc. The filaments generally are filiform, fleshy 
or thin. They are glabrous in all of the species except Celastrus stylosus ssp. stylosus, 
C. aculeatus, C. hirsutus, and C. kusanoi, where they are papillose-tuberculate. The 
filaments usually are longer than the anther, but in Celastrus novoguineensis they 
are shorter. The anthers in the staminate flowers are generally oblong-ovoid, obtuse 
or cordate, but in some species they are distinctly apiculate, i.e., Celastrus pringlei. 
The sterile stamens in the pistillate flowers are small, short, and about 1 mm. long; 
the anthers of these are triangular and cordate, and sometimes they may contain 
sterile pollen grains. 

Fertile and Sterile Pistils: — The fertile pistil usually is superior and free from 
the disc. The ovary usually is subglobose or ovoid, and narrowed into a columnar 
style. The ovary has three cells. Each cell has two ovules in the species of sub- 
genus celastrus while there is only one ovule per cell in the species of subgenus 
racemocelastrus. The stigma is trilobed or the lobes are obscure. In some 
species the lobes are further divided in half, for example, Celastrus orbiculatus. 
The lobes usually are filiform and reflexed, but they are tangentially flattened in 
Celastrus angulatus. The sterile pistil in the staminate flowers usually is small 
and columnar. 

Ovules:— The ovules are anatropous and attached at the base of the ovary by 
a distinct funiculus. The raphe is clearly visible on the seed. In a free-hand sec- 
tion of the ovule the micropyle is sometimes visible towards the base. The ovule 
has the characteristic cup-shaped aril towards its base, which is a definitive char- 
acter of this genus. 

The embryo is erect, thin-foliate and broadly spatulat 

cylindric and stalk-like. 

: 13 . The radicle is 



id or cylindric, and are smooth, or 
The bases of the fruits are obtuse 

column and is stalk-like. The fruits generally have three cells, but two- or four- 
celled fruits are occasionally found on some plants. The £ ™ a J^™J ^ 
six-seeded in most of the species. Celastrus 



all the species of the subgenus racemocelastrus have one developed seed with two 
undeveloped ovules. The fruit is three-valved, dehiscence being locuhcidal. The 
valves are bright yellow or dark brown, slightly transversely wrinkled on the out- 
side, and spreading or reflexed in dehiscence. 

^wTn, A. C. The comparative internal morphology of seeds. Amer. Midi. Nat. )6:520, fig. 2, 

pi. SS- 1946. 

[Vol. 42 



Fig. 1. Chart showing interrelationship of species of Celastrus and their geographical distribution. 



Harris 14 studied the relationship between the number of flowers formed and 
the number of fruits maturing per inflorescence in Celastrus scandens. He also 
studied the size of an inflorescence as measured by the number of flowers it produces 
and the number of seeds per fruit. His conclusion is that there is no correlation 
between these pairs of characters. 

Geographical Distribution 

Celastrus is widely distributed in eastern Asia, Oceania, both Americas and 
Madagascar, between the latitudes of about 40° S. and 47° N. Most of the species 
are found in the subtropical and tropical zones. From the distribution patterns 
of the species, it appears that there are two centers of dispersal: Yunnan-Burma- 
eastern India, and Central America (fig. 2). 

In eastern Asia, the species radiate from the Yunnan-Burma-eastern India 
region. Celastrus paniculatus, one of the two most widely distributed species, 
radiates from that region eastward through Kwangsi, Kwangtung, and Hainan, 
eastward to Formosa; southward through the Malay Peninsula to Indonesia and 
then northward to the Philippines, and westward to India and Ceylon. Celastrus 
bindsii, the other widespread species, extends from the center westward to Sikkim; 
northeastward through Yunnan and Kweichow to Szechuan and Hupeh; eastward 
to Kwangtung, Fukien, Formosa and Bonin Islands, Japan; and southward through 
the Malay peninsula to Indonesia. The other species of that area have much more 
restricted distributions. From the Asiatic center of dispersal, the number of species 


gradually decreases both northward and southward, as is shown on the distribution 

map (fig. 2). 

In North America there is only one species, Celastrus scandens. It is widely 

distributed forming an ellipse from Quebec southward to Virginia, then westward 

to Texas and eastern Wyoming. It is closely related to the Asiatic species of the 

series paniculati. 

The disjunct distribution and floristic relationship of plants of eastern Asia and 

eastern North America recently have been interpreted by Li 15 . He says that this 

familiar disjunct distribution appears to be the remnants of great mesophytic 


period. Geological changes have destroyed and changed 
the floras of many lands so that this mesophytic forest of the Tertiary times in the 


Berry 10 states: "The genus Celastrus Linn, is the largest fossil genus of the 
family. Though its present center of distribution lies in the upper lands of south- 

Harris, J. A. Correlation in the inflorescence of Celastrus scandens. Mo. Bot. Gard. Ann. Rept. 

20'116— 122 1909. 

15 Li, Hui-lin. Floristic relationships between eastern Asia and eastern North America. Trans. 

Amur. Pltifor. S*-. *2:371-429. 1952. § , , r p , 

ia B?rry, E. W. The Lower Eocene flora of southeastern No ta America. U . oc ... jur\. , . • 

P ipti 91:105. 1919. 


[Vol. 42 


eastern Asia and East Indies, its history shows that the ancestral stock was cosmo- 


It is highly 

probable that it originated in America at the dawn of the Upper Cretaceous or 
somewhat earlier." Croizat 17 says that this view cannot be accepted as it is in 
conflict with the distribution data, but he cites no alternative center. All that 
can be said from present distribution patterns is that Celastrus scandens is related 
to the eastern Asiatic center of dispersion rather than to the one in Central 

It is rather difficult to interpret the relationships of the Central and South 
American species. Lcesener 18 has pointed out that the Latin American species show 
relationship with tropical Asiatic species. They closely resemble the Asiatic species 
Celastrus monospermus, C. monospermoides, and C. hindsii, especially in their in- 
florescence patterns, discs, and one-seeded fruits; but differ in the number of ovules 



l^fe ^^f ^lf 

fc^p ** * •■ 













t--r---t- J- 

■ 2 


fcJi^w ij ^g § c.r- 

• - 



Fig. 2. Distribution patterns of Celastrus 

All the maps used in this paper are from Goode's Series 
Press, except the map of the United States, which was kindly 

of Base Maps, University of Chicago 
furnished by Dr. Woodson. 

in each cell of their ovaries and in their sexuality. In the Central and South 

species there is only one ovule in each ovary 

species the ovary 

dioecious while the Latin American species are bisexual. 

Croizat 19 , in his studies of the intercontinental distribution of plants, mentions 
that some Asiatic genera "run a transpacific channel [sic!]" from the Far East to 

'Croizat, Leon. Manual of Phytogeography. 
18 Loes. loc. cit. 20b:l<n. 1942. 

p. 302. 1952 

19 Croizat, Leon. loc. cit. pp. 54, 305, and fig. l6. 1952. 



the New World. He holds the celastraceous genus Perrottetia as a perfect example. 
There are about fifteen species of this genus distributed in Queensland, New Guinea, 
Moluccas, Celebes, Philippines, northern Borneo, Formosa, central China, Hawaii, 
Mexico, Colombia, and Venezuela, and every essential station in the "traject [sic!]" 

is exemplified by one or more species. 

An alternative hypothesis is suggested by the fact that the Latin American 
species of Celastrus constitute a distinct subgenus without representatives in the 
Eastern Hemisphere. These species appear to be closely related to the neighboring 
genus Maytenus, which includes species of both Latin America and Africa, a dis- 
tributional phenomenon which might be explained upon the hypothesis of trans- 
Atlantic continental drift. From a rather dogmatic view of phylogeny, one might 
assume the hermaphroditic flowers of Maytenus (as well as of Gymnosporia) to be 
"primitive" from which the bisexual condition general for the Latin American 
Celastrus might be derived. The North American Celastrus scandens is closely 
related to the species of the Eastern Hemisphere rather than to those of the Western 
Hemisphere, thus suggesting that the North American and Latin American 
Celastrus might have different origins. The two subgenera perhaps therefore can 
be said to have independent origins. 



for the American Celastrus scandens L. and the Asiatic C. orbiculatus Thunb. They 
were able to hybridize these species with interesting results. The hybrid is less 
vigorous than the parents and sparingly fertile; the fruits are smaller than those of 
either parent, and the seeds are not uniform in size. When they back-crossed the 
hybrid to both parents both the F 2 's had leaves closely resembling those of Celas- 
trus scandens. The plants resulting from the back-cross to Celastrus orbiculatus 
yielded one-seeded fruits, with only two exceptions among the 475 fruits pro- 
duced, while the fruits of both parent species contained three to six seeds. Of the 

* 1 "111 

plants resulting from those back-crosses, there is no further record available regard- 

ing the inflorescences, the sexuality of the flowers, the ovule number in each cell 
of the ovary, and the sizes and shapes of the seeds. 

All the Latin American Celastrus species have one-seeded fruits. It would be 
interesting to learn the chromosome number and morphology in these species. Since 
the Latin American species are more or less an uniform and isolated group, cyto- 
logical data might help in relating them to other groups of the genus Celastrus. 

In Celastrus, there appears to be a strong possibility of hybridization taking 
place in nature in certain localities. For example, Celastrus vanioti looks like a 
hybrid between Celastrus angulatus and C. hypoleucus. A specimen collected by 
Wang 76880 from Yunnan, China, seems a hybrid between Celastrus hindsii and 
C. monospermus. 

20 White, O. E., and Bowden, W. N. Oriental and American bittersweet hybrids. Jour. Hered. 

38:125-127. 1947. 


[Vol. 42 


Economic Uses 

There are very few available records of economic uses of Celastrus. Two 
species, Celastrus scandens and C. orbiculatus, are widely cultivated commercially 
for their beautiful yellowish fruits which open at maturity and disclose the crimson 
arillate seeds. 

Balfour 21 states that the bark of Celastrus scandens has emetic and purgative 
properties. The seeds of this species are said 22 to be useful pharmaceutical^. An 
oil is extracted from the seeds of Celastrus paniculatus, which is described as having 
a very hot and bitter taste. It (oleum nigrum) is useful for burning 23 and has 
been employed in India successfully in treating beriberi disease 24 . Burkill 25 reports 
that in Malaya the seeds may be used, externally, in poultices, or in the Philippines, 
internally, as an anti-rheumatic and as a treatment of paralysis. In Java, the leaves 
are used in dysentery. In the Philippines, the sap is given as an antidote in opium- 

Hemsley 26 , quoting from A. Henry's record, states that the leaves and roots of 
Celastrus latifolius Hemsl. (= C. angulatus Maxim.), when dried in the sun and 
pounded to a powder, are efficient in killing insects infesting turnips and other 
vegetables in China. 


This work was completed under the direction and supervision of Dr. Robert 
E. Woodson, Jr. I should like to express my deep gratitude for his kind interest, 
and more particularly for his extraordinary patience and understanding in helping 
me with the problem of my languages. Also, I cannot be sufficiently grateful to 
him for the financial assistance he has obtained for me during the course of my 
study. My thanks are due to Dr. Edgar Anderson, Director of the Missouri 


I also 

acknowledge my gratitude to the Committee on Awards of Fellowships, China 
Institute in America, for the award of the Frank M. Shu Scientific Fellowship for 
the years 1952-53-54, enabling me to undertake this study. My sincere appreciation 
is due to Dr. Rolla M. Tryon, Jr., Dr. Frederick G. Meyer, and Dr. Hui-lin Li for 
suggestions and criticisms during the course of the investigation. I am also in- 

Nell C. Horner for much constructive criticism. I am happy to 



and facilities extended. Finally, I should like to express my thanks to many other 
friends who have directly or indirectly aided and encouraged me in the course of 
this study. 

21 Balfour, J. H. Class Book of Botany, p. 794. 1882. 

"Gray, Asa. The Genera of the Plants of the United States. 2:185. 1849. 

^Lawson, G. in Trans. Bot. Soc. Edinburgh 6:364. 1860. 

1 - V0iS iV J ' ?* ^° Tt l S subu ' banus Calcuttensis. A Catalogue of the Plants which have been 
"IsTu, y , u Indu Com P an y' s botanical garden, Calcutta, etc. p. 166. 1845. 

Burk. 1, I. H. A Dictionary of the Economic Products of the Malay Peninsula. 1:505. 1935. 
Hemsley, W. B. in Jour. Linn. Soc. Bot. 23:124. 1886-88. 




Herbarium specimens used in this study include those of the following herbaria 
or institutions. The abbreviations for them are taken from Lanjouw & Stafleu's 
'Index Herbariorum', part I (Regnum Vegetabile, vol. 2, 2nd. ed., 1954). 

A — Arnold Arboretum, Harvard University, Cambridge, Massachusetts. 
BLAT — Blatter Herbarium, St. Xavier's College, Bombay, India. 
BKL — Herbarium of the Brooklyn Botanic Garden, Brooklyn, New York. 
CAL — Indian Botanic Garden, Howrah, Calcutta, India. 
F — Chicago Natural History Museum, Chicago, Illinois. 
GH — The Gray Herbarium of Harvard University, Cambridge, Massachusetts. 
HK — Herbarium, Gardens Department, Hongkong, China. 
K — The Herbarium, Royal Botanic Gardens, Kew, England. 
L — Rijksherbarium, Leiden, Netherlands. 
MICH — University Herbarium, University of Michigan, Ann Arbor, Michigan. 
MO — Missouri Botanical Garden, St. Louis, Missouri. 
NY — New York Botanical Garden, New York, N. Y. 

P — Museum National d'Histoire Naturelle, Laboratoire de Phanerogamic, Paris, 

PENN — Herbarium of the University of Pennsylvania, Philadelphia, Pennsylvania. 
PH — Academy of Natural Sciences, Philadelphia, Pennsylvania. 

S — Naturhistoriska Riksmuseum, Botanical Department, Stockholm, Sweden. 
SING — Herbarium of the Botanic Gardens, Singapore, Malaya. 
TAI — Herbarium, National Taiwan University, Taipei, Taiwan, China. 

U — Botanical Museum and Herbarium, Utrecht, Netherlands. 
UC — Herbarium of the University of California, Berkeley, California. 
UPS — Institute of Systematic Botany, Botanical Garden and Botanical Museum of the 

University of Uppsala, Uppsala, Sweden. 
US — National Museum, Smithsonian Institution, Washington, D. C. 


Celastrus L. Gen. PL ed. 1. 59. 1737; Sp. PL ed. 1. 196. 1753; Gray, Gen. PL 
U. S. 2:185, pi 170. 1849; Benth. & Hook. Gen. PL 1:364. 1862; Loes. 
in Engl. & Prantl, Nat. Pflanzenfam. 2 Aufl. 20b:131. 1942. (Type species: C. 
scandens L.). 

Euanymoides Isnard, ex Medicus, Philos. Bot. 1:173. 1789 (pro parte). 
Celastrus § Eucelastrus Wight & Am. Prodr. Fl. Ind. Orient. 158. 1834. 
Semarilla Raf. Sylva Tellur. 146. 183 8. (T.: S. bicolor Raf.). 

Scandent shrubs, up to 50 m. tall; branches usually terete, rarely the branchlets 
angular, usually glabrous, rarely pubescent, scarcely to densely lenticellate. Leaves 
extremely variable, elliptic to orbicular, serrate or subentire, rarely finely ciliate- 
serrulate, deciduous, rarely evergreen, alternate, petiolate; stipules small, usually 
laciniate and deciduous. Inflorescences dichasia, paniculiform to racemiform, soli- 
tary, sometimes branched, axillary or terminal, pedunculate or sessile, few- to 
many-flowered. Flowers small, usually unisexual and dioecious, rarely bisexual, 
pedicellate and articulate. Calyx campanulate, persistent on the fruits, 5-lobed; 
petals oblong, obovate-oblong or ovate, glandular-ciliate to erose, or entire, inserted 
under the disc, alternate with the calyx lobes; disc usually membranous and cup- 

[Vol. 42 


shaped, or fleshy and flat, entire or 5-lobed, the lobes alternate with the stamens; 
stamens 5, the filaments glabrous or papillose-tuberculate, arising from the margin 
of the disc proper or attached immediately beneath it, the anthers ovoid or oblong- 
ellipsoid, obtuse or apiculate, dehiscing laterally or extrorsely, dorsifixed, versatile, 
the thecae separated at the lower half or third; ovary superior, subglobose or ovoid, 
free from the disc or rarely slightly confluent with it, 3-celled, each 2-ovuled (sub- 
gen, celastrus) or 1-ovuled (subgen. racemocelastrus) , the ovules arising from 
the base of the ovary, axile, anatropous, with a cup-shaped aril towards its base, 
sessile or with a short funicle, the style usually columnar, the stigma usually 3- 
lobed or the lobes obscure, rarely each bifid. Fruit a capsule, usually subglobose, 
rarely cylindric, tipped by the persistent style, 3-celled, dehiscence loculicidal, 3- 
valved, the valves terminating in part of the style, 1- to 6-seeded; seeds enclosed 
in a fleshy crimson aril, the areolae distinct or obscure, the albumen copious, the 
embryo erect, thin-foliate and broadly spatulate. 


A. Flowers unisexual and dioecious; ovary 3-celled, each cell 2-ovuled; fruits 3- to 6- 
seeded, or in some species 1 -seeded with 5 distinctly undeveloped ovules. Asia, Oceania, 

Madagascar, and North America Subgen. I. Celastrus 

B. Inflorescences usually terminal only, or, if in the axils of the uppermost leaves as well, 

without accompanying vegetative buds Ser. 1. paniculati 

BB. Inflorescences both terminal and axillary, or axillary only, but the axillary always 
with accompanying vegetative buds (except in C. monospermus, C. monospermoides, 

and C. hindsii) Ser. 2. axillares 

A A. Flowers bisexual; ovary 3-celled, each cell 1-ovuled; fruits 1 -seeded with 2 distinctly 

undeveloped ovules. Central and South America Subgen. II. Racemocelastrus 

Subgenus I. Celastrus 
Series 1. paniculati Rehd. & Wils. in Sarg. PL Wils. 2:354. 1915. 


A. Calyx lobes imbricate, broader than long. 

B. Inflorescences usually at least thrice compound; petioles usually more than 10 mm. 

C. Anthers of the male flowers truncate at the apex, with distinct filaments about 1 
mm. long; pistil of the female flower about 2.5 mm. long, the style half as long 
as the ovary. India, Burma, Siam, Indo-China, China, Malaya, Indonesia, and 

the Philippines 1. C. paniculatus 

CC. Anthers of the male flowers apiculate, sessile or subsessile; pistil of the female 
flower about 1.5 mm. long, the style about one-fourth as long as the ovary. New 

Guinea 2. C. novoguineenm 

BB. Inflorescences once or twice compound; petiole less than 10 mm. long. 

D. Anthers of the male flowers not apiculate or obscurely so; stigmata distinctly 
3-lobed; disc-lobes truncate; inflorescences usually 3-9 cm. long. 
E. Fruiting pedicels 1-3 mm. long; leaves firmly membranous. Eastern Australia, 

Papua, and New Caledonia 3. Q. subspicatus 

EE. Fruiting pedicels 4-12 mm. long; leaves thinly membranous. Fiji 4. C. richii 

DD. Anthers of the male flowers apiculate; stigmata discoid or slightly 3-lobed; disc- 
lobes cuspidate; inflorescences 1-3 cm. long. Madagascar 5. C. madagascariensh 

AA. Calyx lobes open, longer than broad. 

F. Young branchlets angular; leaves orbicular or broadly ovate, the petioles of 
upper leaves usually more than 2 cm. long; inflorescences usually more than 
10 cm. long, dense, many-flowered; disc fleshy, flat; fruits subsessile. 
China 6. C. angulatus 



FF. Young branchlets terete; leaves elliptic or broadly elliptic, the petioles of 
upper leaves usually less than 1.6 cm. long; inflorescences usually less than 
8 cm. long, lax, relatively few-flowered; disc membranous, cup-shaped; fruits 
pedicellate, the pedicels 2—5 mm. long. North America 7. C. scandens 

1., 1:1125. 1797 (as paniculata ) ; Lawson 
in Hook. Fl. Brit. Ind. 1:617. 1875, ex char. 

Scandent shrubs up to 10 m. tall; branches terete, glabrous, brown, the young 
branchlets terete, usually pubescent, densely lenticellate, the lenticels elliptic, some- 
times elevated; axillary buds deltoid to orbicular, about 1.5 mm. long. Leaves 
elliptic, obovate, suborbicular, broadly ovate, ovate-oblong to oblong, the apex 
abruptly acute, obtuse or rarely emarginate, the base cuneate, broadly acute to 
obtuse, the margins serrate or remotely crenate, 5-15 cm. long, 2.5-6.0 cm. wide, 
firmly membranous, glabrous, sometimes pubescent on the veins below, the primary 
lateral veins usually 5-7 pairs, elevated on both surfaces, curved toward the apex, 
the veinlets distinct to slightly elevated below, immersed but visible above; stipules 
laciniate, about 1 mm. long; petioles 0.5-1.5 cm. long. Inflorescences terminal, 
paniculiform, thrice- to multi-compound, spreading, usually 5-10 cm. long, rarely 
up to 20 cm. long, pedunculate, the peduncles glabrous or puberulous, the primary 
peduncles usually 6—10 mm. long; flowers dioecious, pale or yellowish green, the 

mm. long, the articulation at the basal part of the stalk. Male 
flowers: calyx lobes semi-orbicular, imbricate, ciliate, about 0.7 mm. long and 1.5 
mm. wide; petals oblong or obovate-oblong, obtuse, entire, about 3 mm. long and 
1.5 mm. wide, with parallel veins; disc cup-shaped, the lobes inconspicuous, slightly 



filaments subulate, flat, glabrous, the anthers ovoid, obtuse, cordate; sterile pistil 
columnar, about 1.3 mm. long. Female flowers: calyx lobes, petals, and disc as in 
the male; sterile stamens 1.3 mm. long; pistil 2.5 mm. long, the ovary globose, the 
style columnar, the stigmata 3-lobed, each deeply bifid, slender. Fruits subglobose, 
the valves broadly elliptic, about 6-9 mm. long and 5-8 mm. wide, 3- to 6-seeded; 


Chiefly in thickets, at altitudes from 200 to 1,800 m.; widely distributed in 
India, Burma, Siam, Indo-China, China, Malaya, Indonesia, and the Philippines; 
flowering from April to June. 


A. Leaves orbicular to broadly ovate. 

B. Leaves orbicular, suborbicular, or obovate. India, Ceylon, Burma, Indo-China and 

southwestern China la - SS P- paniculatus 

BB. Leaves broadly ovate. Philippines and Indonesia lb. $S P- ^erratus 

AA. Leaves elliptic to elliptic-oblong. India, Burma, Siam, Malay, Indonesia, Palawan Isl. 

(Philippines), and China lc. «P- multifloms 

la. Celastrus paniculatus ssp. PANICULATUS. 

Celastrus nutans Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind., ed. Carey & Wall. 

2:390. 1824, ex char. 
Celastrus rothiana Roem. & Schult. Syst. 5:423. 1819; DC. Prod. 2:8. 1825, ex char. 

(T.: Heyne s.n.). 


[Vol. 42 


Ceanothus paniculatus Heyne ex Roth, Nov. PL Sp. 154. 1821, ex char. 

Scutia ? paniculata G. Don, Gen. Syst. 2:34. 1832, ex char. 

Celastrus pubescens Wall. Cat. 4303. 1831, nom. nud. 

Celastrus metzianus Turcz. in Bull. Soc. Nat. Mosc. 31:448. 1858. (T.: Metz 1549, GH!). 

Celastrus paniculatus var. pubescens Kurz ex Prain, in Jour. Asiat. Soc. Bengal 73:196. 

1904, nom. nud. 
Celastrus paniculatus var. andamanica Kurz ex Prain, loc. cit. 73:196. 1904, nom nud. 

Fig. 3. Celastrus paniculatus ssp. paniculatus 

Burma: Fort Stedman, Abdul Khalil s.n., 1894 (CAL); Dr. King's collector 454 
(CAL), 479 (L); Cokteih Shaub, Meebold 8055 (S); Maymyo Hill, Badal Khan 36, 259 
(CAL); Maymyo Plateau, Lace 3130, 3280 (CAL); Pegu, Eyre s.n., 1887 (CAL), Kurz 
1926 (CAL); shan hills: Keloh, Collett 588 (CAL), Pwehla, Collett 672 (CAL); 
Shwebo, Abdul Huk 62 (CAL); southern shan state: Laikow, Abdul Khalil s.n., 
1893 (CAL), s.n., 1894 (CAL), Taungyi, Abdul Khalil s.n., 1893 (CAL), s.n., 1894 
(CAL, L, UPS) ; Yeu, Cole s. n., 1890 (CAL). 

Ceylon: Rewamb, Watson 176 (CAL); without precise locality, Thwaites 

China: yunnan: Che-li, Wang 75551, 76335 (A) ; Fo-hai Hsien, Wang 74570, 74650 
(A); Nan-chiao, Wang 75024 (A); Shung-kiang Hsien, Wang 72996 (A); Lang-tsang 
Hsien, Wang 76644 (A); Szemeo, Henry II 572 (US). 

India: assam: Khasia Hills, 1,000-3,000 ft., Hooker tf Thompson s.n. (CAL, L, S, 
U), Duphla Hills, Radhu, Pokri, Lister s. n. (CAL) ; Andamann's, Dr. King's collector s.n., 
Aug. 2, 1891 (CAL), Dr. Train's collector 70, 78 (CAL). bengal: Lower Bengal, Kurz 
s.n. (CAL); Darjelling, Reyang, Osmaston s.n., Nov. 25, 1903 ((CAL), Manbhum, 
Campbell s.n. (CAL); Maymbhing, Baripada, Hooper 38801 (CAL), Orissa, Ball s.n., 
April 26, 1896 (CAL); Pachete, Kurz s.n. (CAL). bihar state: Chota-Nagpur, 
Parasnath, Clarke 24857 (CAL); Chota Nagpur, Palamow, in Kumandi Reserve, Gamble 
8780 (CAL); Golah, Hazaribagh, Chota Nagpur, Train s.n., Nov. 29, 1891 (CAL); 
Parasnath, Kurz s.n. (CAL); Rajmahal Hills near Sahebganj, Kurz s.n., May 1867 
(CAL); Topechana, Kurz 65 (CAL). Bombay state: vie. Concan, Stocks s.n. (CAL, 
L, S. U, UC, UPS); Khandala, Santapau S. J. 12707 (MO): Poona. Deccan. Woodrow 



S.n., 1882 (CAL); Dharasu, Dudgeon tf Kenoyer 237 (MO); Bhawani, in montibus 
Nilagincis, Metz 1549 (type of Celastrus metzianus Turcz., UPS). Himalaya: Bhim-Tal, 
Kumao, Strachey tf Winterbottom 2 (CAL); Kalidungi, Thomson 740 (CAL); Lachin- 
wala, Nai Sanon Samnasena 34 (CAL); Pattludon, Brandts 2032 (CAL); Raipur, Clarke 
23681* (CAL); Kalsi & Asarori, Amar Nath 24 (A); hills northwest of Kendat, Prazer 
273 (CAL); Madhyapradesh, King s. n. (CAL). madras state: Ganjam, Pralugaon, 
Chilka Lake, Hooper 39537 (CAL); Godairi, Chodavarum, Ramaswami 1 490 (CAL); 
Naikeneri, Palmaner-chittur, Fischer 4739 (CAL); Rambba, ti ar ay an asw ami 105 (CAL), 
madhya pradesh: Gwalior, Maries 22 (CAL); Khandwa, Duthie 8211 (CAL). Mysore: 
Somanathapur, Barber 6846 (CAL) ; Shimoga, Barber 7018 (CAL) ; Bettiah, near the Nepal 
frontier, Hieronymus 443 (CAL). Hyderabad: Jalna, Peninsular India, Rodriguez 202$ 
(CAL). Punjab: Bhadwar, Kangra, Koelz 4266 (MICH, US), 4377 (MICH, UC, US); 
Gurdaspur Distr., Bis Ram 309 (S) ; Karnal DJstr., Drummond 21651, 21691 (UC); 
Kasauli Hills, Drummond 21515 (UC) ; Kasansi, Drummond 21 516 (UC) ; Lachhiwala, 
Revoti Mohon Mukherjee 32 (US), travancore cochin state: Shencotah to Aryan 
Karn, Calder & Ramaswami 654 (CAL); Nokara, Rama Rao 1 588 (CAL). sikkim: 
Garubathen, Dr. Pram's collector s. n., 1,500 ft. alt., 1900 (CAL); Rungit, Clarke 26330 
(CAL) ; without precise locality, alt. 2,000-4,000 ft., Hooker s.n. (CAL, L). rajasthan: 
Abu, King s.n. (CAL), Stocks 231 (CAL). uttar pradesh: Dehra Dun, Pradhan 34 
(S) ; Piujour & Dehra Dun, Anrudh Singh 25 (S), Vicary (?) 175 (CAL) ; Upper Gangetic 
plain, Harsukh 21400 (UC), Dehra Dun, Mackinnon s. n. (CAL); Singh 43 (UC). 
Indo-China: Cambodia, Pierre 895 (L). 

lb. Celastrus paniculatus ssp. serratus (Blanco) Ding Hou, stat. nov. 

Diosma serrata Blanco, FI. Filip. 168. 1837; ed. 2. 119. 1845, ex char. 

Celastrus polybotrys Turcz. in Bull. Soc. Nat. Mosc. 31:449. 1858. (T.: Cuming 1321). 

Indonesia: Celebes: Prov. Minahassa, Koorders 19634P (L) ; Makale, Kjellberg 2907 

(S) ; Paloe, Curran 3415 (A). Java: Prov. Bawean, Coert 3758 (L) ; Moenoeng, de Voogd 

682 (A, L) ; Poeger, Besochi, Koorders 30178P (L, UC) ; Kawi, Mousset 1 1 II (L). timor 

Nonboumn, Teysmann 8731 (L); Soe, de Voogd 2338 (A, L). 

Philippines: bohol: Ramos s.n., Aug.-Oct. 1923 (A, UC). luzon: Prov. Bataan 

7499 (SING) 


Prov. Bulacan, Ramos 346 (A, F, L, MO, US); Cagayan, Cuming 1324 (L, NY, UPS); 
Prov. Cavite, Ramos ef Deroy s.n., April-May 1915 (L, US); Prov. Nueva Viscaya, Mc- 
Gregor s. n., March-April, 1912 (MO, S) ; Prov. Pangasinan, Otanes s. n., April- June 1914 
(A, SING, US); Rizal, Ramos 9 (US); Ahern's collector s.n., March 1905 (F, SING, 
US) ; Prov. Tayabas, Alcasid & Edano 4669 (A) ; Prov. Zambales, Pox 4864 (A) ; without 
precise locality, Cuming 1209 (L, MO, UPS). Mindanao: Dist. Cotabato, Robinson s. n., 
June 1910 (F) ; Pagpawan Sitio, Davao Prov., Edano s.n. (L); Pandarochan, Merrill 944 

lc. Celastrus paniculatus ssp. multiflorus (Roxb.) Ding Hou, stat. nov. 

Celastrus multiflorus Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind. ed. Carey & Wall. 

2:389, 1824; Fl. Ind. ed. Carey 1:622. 1832, ex char., non Lamarck. 
Celastrus dependens Wall, in Roxb. Fl. Ind. ed. Carey & Wall. 2:389. 1824, in obs. 

Celastrus multiflorus Roxb., nom. nud. 
Alsodeia glabra Burgdk. in Junghuhn, Pi. Jungh. 122. 1881-85. (T.: Junghuhn s. n. f LI). 
Celastrus laotica Pitard, Fl. Gen. l'Indo-Chm. 1:891. 1912. (T.: Pierre 2794, PI). 
Euonymus euphlebiphyllus Hayata, Icon. PI. Formos. 5:15. 1915, ex char. (T.: Fujii s.n.). 
Celastrus euphlebiphyllus (Hay.) Makino & Nemoto, Fl. Jap. ed. 2. 597. 1931. 
Celastrus euphlebiphyllus (Hay.) Kanehira, Formos. Trees, ed. 2. 383, /. 340. 1936, nom. 


[Vol. 42 


Burma: Chin Hills, Prazer 236 (CAL) ; Chindwin, near Tummoc, Prazer 167 (CAL) ; 
Kachin Hill, Sbaik Mokim s. n. y 1897 (CAL) ; Koni, Prazer s. «., May 1888 (CAL) ; Luay 
Wyo, Rodger 198 (CAL); Maymyo Hill, Badal Khan II (CAL); Myitkyina, Rogers 864 
(CAL) ; Pinmona, Abdul Huk s. n., Aug. 26, 1890 (CAL) ; southern Shan State, Taungyi, 
Abdul Khalil s.n. y 1894 (CAL); Tavoy, Oomaigoi, Sbaik Mokim 266 (CAL, L, U) ; 
Younghue, Carter 466 (CAL); without precise locality, Prazer 54 (CAL), 236 (L). 

China: Hainan: Ka La, McClure Q182 (A, MO); Kan-en Hsien, Lau 3658 (A, S) ; 
Ngai Hsien, Lau 283 (A, MICH, MO, UC, US) ; Yaichow, Liang 63042 (A), How 70573 
(A, US). Taiwan: Kurau, Kosyun, Uo s. n. (photo, A), yunnan: Che-li, Wang 75551, 
77833> 7956o (A); Cheng-kango, Monghui, Yu 16787 (A); Fo-Hai Hsien, 74966 (A); 
Jenn-yeh Hsien, Wang 80588, 80194 (A); Shunning, Yu l62Q2 (A); Szemao, Henry 

I2I22 (MO), I2I22*, 12122^ (US), 12572* (A), IIQ72 A (US), II993 (A, MO), 



Indo-China: Prov. Sonla, 4049 (A, UC) ; Bao-chiang, Prov. Bien-hoa, Pierre 2794 
(lectotype of Celastrus laotica Pitard, P) ; Phuoc-than, Tborel s. n. y 1862-66 (paratype of 
C. laotica Pitard, P) ; Laos, Phon-thane, Spire 289 (paratype of C. laotica Pitard, P). 

Indonesia: Celebes: Minahassa, Lam 2466 (L). java: Bantam, Kuhl & v. Hasselt 
18 (L) ; Madiun, Ngebel, Koorders 2980B (L) ; Parve-sosa, Backer 37252 (L) ; Pehalongan, 
Koorders 27372B (L); Soerabaja, Dorgelo 2290 (L). Sumatra: without precise locality, 
Junghubn s. n. (type of Alsodeia glabra Burgdk., L). 

Malay Peninsula: Pahang: Kwala Sunbelung, Mat s.n., 1893 (SING); Sea Pardens, 

Ridley s.n., 1892 (SING) 

lerrill 9311 (A, F, L, MO, SING, US). 

SING); Bitsenuloke, Groff 6132 (UC); Kin Sayo, 


441 (L, SING) ; without precise locality 

The wide distribution and the variable leaf forms of this species have been the 
cause of the many synonyms. Because of the inadequacy of flowering material, I 
selected the middle leaf of a branch of current season's or year's growth for com- 
parative statistical study, following Woodson's 27 method of measuring leaf char- 
acters in Asclepias. A study of the herbarium material according to the geo- 
graphical distribution and measurable differences among the populations has demon- 
strated that this species includes three distinct populations, which I am treating as 
three subspecies (pi. 34, upper diagram). 

Celastrus paniculatus ssp. paniculatus has leaves which are usually suborbicular 
or broadly obovate. It is chiefly distributed in India, but it extends to Burma, 



rerlaps ssp. paniculatus. It 
is also distributed in Hainan, Formosa, Siam, Indo-China, Malaya, Indonesia, and 
Palawan IsL, in the westernmost Philippines. The leaves of ssp. serratus are usually 
broadly ovate. This subspecies is chiefly distributed in the Philippines, and is less 
common in Java where it overlaps ssp. multtflorus. 

27 Woodson, R. E., Jr. Some dynamics of leaf variation in Asclepias tuberosa. Ann. Mo. Bot. 
Gard. 34:353-432. 1947. 

Ann. Mo. Bot. Gard., Vol. 42, 1955 

Plate 34 




















o o 














1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.3 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0 


Frequency distribution of three subspecies of Celastrus paniculatus: large dots, up. paniculatus; 
small dots, ssp. serratus; circles, ssp. multiflorus. Explanation in the text. 

[Vol. 42 


I have not seen any specimen of C. paniculata collected in Borneo, but I think 
that it may be found there from the distribution pattern. It is interesting to note 
from the geographical distribution of these three subspecies that the specimens 
collected in Palawan Island and Formosa belong to the ssp. multiflorus instead of 
the Philippine ssp. serratus. Geologically, Palawan Island and Formosa belong to 
the Malaysian continent. 

The leaves borne on a branch of current season's or year's growth are slightly 
variable. If a leaf picked at random from a plant of one subspecies is compared 
with that of another subspecies the same size and form can be found. Following 
Anderson's 28 method, I have traced the leaf forms on a current season's or year's 
branch for each subspecies, and the differences in leaf patterns were easily observed. 

2. Celastrus novoguineensis Merr. & Perry, in Jour. Arn. Arb. 22:260. 1941. 
(T.: Clemens 5152, A!). 
Scandent shrubs; branchlets terete, glabrous, brown to reddish-brown; the 

lenticels numerous, elliptic, elevated, white; axillary buds conoid, obtuse, about 2 
mm. long. Leaves oblong, ovate to elliptic, the apex acute or obtuse, the base 
obtuse or rotund, the margins minutely and remotely serrulate, 12-16 cm. long, 
5.0—7.5 cm. wide, firmly membranous, glabrous, the primary veins 6 pairs, elevated 
below, distinct, plane or slightly elevated above, the veinlets obscure below, plainly 
visible above; stipules linear, about 2.5 mm. long; petioles 1.0-1.5 cm. long. 
Inflorescences terminal, paniculiform, usually thrice compound, about 15-20 cm. 
long, the floriferous branchlets divaricate, the peduncles glabrous, the primary 
peduncles about 6—10 mm. long; flowers green, dioecious, the pedicels about 1 mm. 
long, the articulation varying from the lower to upper half of the stalk. Male 
flowers: calyx lobes imbricate, subreniform, slightly erose, about 1 mm. long; 
petals oblong, rotund, subentire, scarious-marginate, about 2.5 mm. long and 1.5 
mm. wide, subfleshy, brownish-punctate; disc cup-shaped, membranous, the lobes 
obscure, truncate or mucronate; stamens subsessile or with very short flat filaments, 
the anthers narrow-ovoid, obtuse, apiculate, brownish-dotted; sterile pistil 1 mm. 
long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens 
1 mm. long; pistil about 1.4 mm. long, the ovary subglobose, wider than long, the 
style obsolete, the stigmata 3-lobed, each bifid. Fruits subglobose, the valves broadly 
ovate, about 9—12 mm. long and 7-10 mm. wide, 3- to 6-seeded; seeds ellipsoid, 
5—8 mm. long and 2.5—3.5 mm. wide, reddish-brown, smooth. 

In thickets, at altitudes 1,220-1,830 m.; New Guinea, flowering from February 
to September. 



(A); Yoangen (Yunzaing), /. & M. S. Clemens 3523, 6606, 7230 (A), papua: Lala 
River, Carr 15608 (A). Netherlands new guinea: Meervlakte, Motor bivouac, f. 
Leeuwen 1 1 008 (L). 

28 Anderson, E. Concordant versus discordant variation in relation to introgression. Evolution 
5:133-141. 1951. 




Map 1. 



[Vol. 42 



This is an endemic species of New Guinea. It is distinctive and can be dis- 
tinguished from other species by the thrice compound aggregate dichasia and main 
floriferous axis with branchlets extending out at nearly right angles. The stamens 
in the male flowers usually are sessile or with very short, flat filaments. The pistil 
in the female flower has a very short style. 

3. Celastrus subspicatus Hook. Icon. PL 5:t. 482. 1842. (T.: Collector un- 
known, s. n. y cultivated at Kew, K!). 

Celastrus australis Harv. & Muell. in Trans. Phil. Soc. Vict. 1:41. 1855; Benth. Fl. 

Austral. 1:398. 1863, ex char. (T.: Mueller s.n.). 
Celastrus papuana Warb. in Engl. Bot. Jahrb. 13:366. 1891, ex char. (T.: Warburg s.n.). 
Celastrus paniculatus Willd. var. Balansae Loes. in Engl. Bot. Jahrb. 39:160. 1906, ex 

char. (T.: Balansa 3029). 

Scandent shrubs; branchlets terete, brownish-pubescent, the lenticels elliptic, 
distinct, and white; axillary buds deltoid, spiniform, about 1-2 mm. long, the 
outermost scales acuminate. Leaves elliptic, ovate, suborbiculate, or ovate- 
lanceolate, the apex acuminate, the base obtuse to acute, the margins subentire to 
remotely serrulate, 4-12 cm. long, 1-6 cm. wide, membranous, glabrous, the 
primary lateral veins 5—8 pairs, the veins and veinlets slightly elevated on both 
surfaces; stipules linear, about 2.5 mm. long; petioles usually 4—7 mm. long. In- 
florescences terminal, once compound, rarely twice compound in the male, 2-9 
cm. long, occasional in the axils of the uppermost leaves, the peduncles usually 
brownish-puberulous ; flowers white, the pedicels about 3 mm. long, the articula- 
tion toward the base of the stalk. Male flowers (young) : anthers ovoid, obtuse. 
Female flowers: calyx lobes imbricate, orbicular, broad, cilia te, about 1 mm. long; 
petals ovate-oblong, rotund, slightly erose, about 2.5 mm. long and 1.0-1.5 mm. 
wide; disc cup-shaped, the lobes obscure, truncate; sterile stamens 0.5 mm. long, 
the filaments very short; pistil ovoid, 2 mm. long, the style short, the stigmata 
3-lobed, spread. Fruits subglobose, the valves broadly elliptic, about 5—10 mm. 
long and 5—8 mm. wide, 3- to 6-seeded; seeds ellipsoid, 5 mm. long and 2.5 mm. 
wide, reddish-brown, the areolae obscure. 

In the thickets; Australia, New Caledonia, and Papua. 

Australia: Queensland: Cedar Beenleigh, White s.n., May 1920 (SING); head of 
Burnett R., Miiller s.n. (GH); Benarkin, Cameron s.n., April 1924 (A); Mt. Glorious, 



(UC); without definite locality: Strathdickie 1 124. (GH), Clemens 43683 (A), new 
south wales: Success Hill, Caley s.n., 1807 (A, UC, US); Tweed River, Moore s.n. 
(GH) ; cultivated at Kew, collector unknown s. n. (holotype of Celastrus subspicatus 

Hook., K). 

New Caledonia: Balansa 1 8 70 (A); De planch 92 (A). 

New Guinea: papua: Kanosia, Carr 11260 (A, L). 

Celastrus subspicatus was published by W. J. Hooker, based on a plant long 
cultivated in the Royal Botanic Gardens at Kew. Its history and habitat are 
unknown. In the type specimen, borrowed from Kew Gardens, the leaf forms, 
texture, short petioles, distinct and elevated lenticels, acuminate outermost 






larity leads me to believe that Celastrus subspicatus of Kew Gardens was grown 
from seeds obtained from Australia. 

Specimens collected from New Caledonia and lowland Papua, New Guinea, are 
morphologically similar and geographically related to the Australian ones except 
in the leaf shape. Meanwhile, not enough specimens are available for a population 
study to find out whether they really differ from the Australian species and whether 
they should be put into subspecific or varietal category, so they are treated tenta- 
tively as one species. 

4. Celastrus richii A. Gray, in Wilkes U. S. Expl. Exped. (Bot. Phanerogam., 

pt. 1):289. 1854. (T.: Wilkes Exped, s.n., GH!). 

Scandent shrubs; branchlets terete, glabrous, light to dark brown, the lenticels 
distinct and numerous, white and elliptic, rarely lacking on the young branchlets; 
axillary buds small, conic, about 1 mm. long. Leaves elliptic, the apex and the 
base acute, the margins subentire or remotely serrulate, thinly membranous, shining 
above in dry condition, glabrous, the primary veins 5-7 pairs, elevated on both 
surfaces, the veinlets distinct on both surfaces, densely reticulate; stipules laciniate, 
filiform, about 1 mm. long; petioles 2-9 mm. long. Inflorescences terminal, once 
or twice compound, 2—7 cm. long, the peduncles glabrous, the primary peduncles 
3.5-5.5 mm. long; flowers white, the pedicels 1.0-2.5 mm. long, accrescent, up to 
15 mm. long on the fruits, the articulation toward the base of the stalk. Male 
flowers: calyx lobes imbricate, subreniform or semi-orbicular, about 1 mm. long, 
ciliate; petals oblong, entire, 3 mm. long and 2 mm, wide; disc concave, mem- 
branous, the lobes obscure, mucronate, the margins confluent with the base of the 
filaments; stamens about 2 mm. long, the filaments short, flat, broad at the base, 
gradually attenuated toward the top, glabrous, the anthers ovoid, slightly cordate 
and apiculate; sterile pistil conoid, 1.5 mm. long. Female flowers: calyx lobes and 
petals as in the male; disc cup-shaped, membranous, the lobes subreniform; sterile 
stamens 0.8 mm. long; pistil 2 mm. long, the ovary subglobose, the style short, the 
stigmata 3-lobed, each bifid. Fruits subglobose, the valves suborbicular or obovate, 
about 9 mm. long and 7-8 mm, wide; seeds ellipsoid, about 4.5 mm. long and 3 

mm. wide, reddish-brown, distinctly areolate. 

In sunny forest; at altitudes 30-900 m.; Fiji Isl.; flowering from December to 

Fiji: fulanga: Smith 1122 (GH, NY, S, UC, US), vm levu: Lautoka, Greenwood 
OOO (A) ; Mba, Smith 4301 (A, US), 5827, 6085 (A) ; Tholo, Degener 14798 (NY, US), 

14.940, 14973 (A, NY, US), 14978 (A, MICH, NY), 13050, 153*5 (A, MICH, NY, S, 
US) ; Gillespie 4182 (GH, NY, UC) ; Nandi, Degener 15330 (A, NY, S, US) ; Ra, Degener 
15352 (A, NY), 15424 (A, MICH, NY, S, US), 15427 (A, NY), without precise 

locality: Home 1135 (GH); Mrs. Parham 3 (GH); Wilkes Exped. s.n., 1838-42 (GH, 
holotype) . 


[Vol. 42 

This is an endemic species of Fiji, Its geographical distribution and accrescent 
pedicels make this species easily separable from other species. A specimen collected 
by Degener (15330) in an "arid patch of forest along coast" near Nandi has 
obovoid fruits while all others have subgloboid ones. 

5. Celastrus madagascariensis Loes. in Notiz. Bot. Gart. Berlin 13:215. 1936, 

ex char. (T.: Perrier de la Bat hie 2067). 

Scandent shrubs; branches terete, glabrescent, brown, the lenticels elliptic or 
orbicular, slightly elevated; axillary buds globose to deltoid, about 1 mm. long. 
Leaves usually elliptic to ovate-oblong, the apex acute to shortly acuminate, the 
base cuneate or obtuse, the margins crenate-serrulate, 4-7 cm. long, 2.5—3.5 cm. 
wide, chartaceous to firmly membranous, glabrous, the primary lateral veins 5-7 
pairs, elevated on both surfaces, curved toward the apex, the veinlets distinct and 
slightly elevated on both surfaces; stipules laciniate, about 2 mm. long; petioles 
5-11 mm. long. Inflorescences terminal, paniculiform, once compound, 1-3 cm. 
long, the peduncle puberulous, the primary peduncles 2-3 mm. long; flowers 
dioecious, the pedicels about 2.5 mm. long, the articulation toward the base of the 
stalk. Male flowers: calyx lobes imbricate, ovate, obtuse, slightly erose, about 1 
mm. long; petals elliptic to ovate, obtuse and subentire, about 3 mm. long and 
2 mm. wide; disc membranous, the lobes inconspicuous, cuspidate; stamens arising 
from the margin of the disc, about 3 mm. long, the filaments shorter than the 
anthers, glabrous, the anthers ellipsoid, distinctly apiculate; sterile pistil columnar, 
about 1 mm. long. Female flowers: calyx lobes, petals and disc as in the male; 
sterile stamens about 1 mm. long; pistil about 2.5 mm. long, the ovary globose, 
the style shortly columnar, the stigma discoid. Fruit unknown. 

In forests; at altitudes 100-1,500 m.; endemic in Madagascar; flowering from 
October to November. 

Madagascar: Valley of the Mandrare, d'Alleizette 1288 (L) ; Basaltic near Betafo, 
d'Alleizette s.n., Sept. 1905 (L). 

This species quite clearly belongs to the series paniculati by its distinct 
terminal paniculiform inflorescences. It is closely related to the Australian 
Celastrus subspicatus, rather than to the Asiatic C. paniculatus, by its short and 
once compound inflorescences. It can be distinguished from those two species by 
its apiculate anthers, cuspidate disc-lobes, and discoid stigma. Loesener has pointed 
out that the Celastrus species of Madagascar enumerated by Drake 29 are Gymno- 
sporia. The excellent illustrations reveal that Loesener's view is quite right. Thus 
Celastrus madagascariensis is the only remaining Celastrus species described from 


Drake del Castillo, M. E., in Grandidier's Hist., Phys., Nat. et Pol. Madag. 36 (Hist. Nat. Pi. 
Atlas y):ph. 280, 280*, 280*. 1896. 




Fig. 4. Celastrus angulatus Maxim 

6. Celastrus angulatus Maxim, in Bull. Acad. Sci. St. Petersb. Ill, 27:455. 
1881 (as angulata); in Mel. Biol. Acad. St. Petersb. 11:199. 1881, ex char. 
(T.: Piaseski s.n.). 

Celastrus latifolim Hemsl. in Jour. Linn. Soc. 23:123. 1886; Oliver, in Hook. Icon. Pi. 
t. 2206. 1894, ex char. & ill. (T.: Henry j. /*.). 

Scandent shrubs up to 10 m. tall; branches subterete, the branchlets usually 
angular, glabrous, both reddish to dark brown, shining, densely lenticellate, the 
lenticels small, orbicular or ovate, white; axillary buds short-ovoid, about 2-5 mm. 
long. Leaves broadly elliptic, broadly ovate to suborbicular, the apex abruptly 




primary lateral veins usually 6-7 pairs, prominently elevated below, plane or 
slightly elevated above, the secondary lateral veins slightly elevated below, plane 
and visible above, nearly parallel to one another; stipules filiform, tufted; petioles 
0.6-3.0 cm. long. Inflorescences terminal, densely paniculiform, up to 20 





[Vol. 42 

tion just beneath the flower. Male flowers: calyx lobes open, ovate or oblong-ovate, 
obtuse, subentire, about 1.5 mm. long; petals elliptic-oblong, obtuse, ciliate to 
slightly erose, reflexed, about 2.5 mm. long and 1.2 mm. wide; disc subfleshy, 
nearly flat, plate-like, the lobes usually deprcssed-subquadrate; stamens attached 
just under the margin of the disc, about 3 mm. long, the filaments filiform, 
glabrous, the anthers ovoid, obtuse, apiculate; sterile pistil ovoid, about 1 mm. long. 
Female flowers: calyx lobes, petals and disc as in the male; sterile stamens about 
1 mm. long; pistil about 3—4 mm. long, the ovary subglobose, the style distinctly 
columnar, the stigmata 3-lobed, each bifid, recurved. Fruits subglobose, the valves 
broadly elliptic to suborbicular, about 6—9 mm. long and 5-9 mm. wide, 3- to 6- 
seeded; seeds ellipsoid, about 3-5 mm. long and 1.5-3.0 mm. wide, brownish-red, 
shining, the areolae usually obscure, rarely distinct. 

In open places, light woods and thickets; at altitudes of 400-3,600 m.; China; 
flowering from May to June. 

China: honan: Hia Hsien, Hers 1831 (A); Hwei Hsien, Hers 772 (A); Lushih, 
Hiung-eul Shan, Hers 861 (A); Sung Hsien, Hers 1288, 1 314 (A); Teng-feng Hsien, 
Hers 26Q (A) ; Tsi-yuan Hsien, Hers 1712 (A) ; Yungning, Hers 417 (A), hupei: Chang- 
yang, Wilson 753A (A); Enshih Hsien, Ho-ch'ang Chow 200 5 (A, NY); Fang Hsien, 
Wilson 364% 7 S3 (A, F, GH, MO, US) ; Ho-long-teng, Silvestri 5099 (A); Ichang, Chun 
tf Chien 8050 (UC), Henry 2084, 3405, 3883 (GH), 5925 (A, GH, NY), Wilson 364 
(A, F, MO, US); Ma-pan-scian, Silvestri 1336 (A); Nanto, Wilson 783 (A); Paokan, 
Wilson 467 (A, NY); Patung Hsien, Ho-ch'ang Chow 495, 597 (A, NY); Sin-yeh-sie, 
Chun 3724 (A), 4066 (US) ; Tzu-kwei Hsien, Ho-ch'ang Chow 405 (A, NY), kansu: 
Tienschui, Fenzel 2817 (A), kwangsi: Ch'uan Hsien, Tsang 27666 (US), kwangtung: 
Loh-ch'ang, Tsang 20813 (A, MO, NY, SING, UC), Tso 21152 (NY), kweichow: 
Kiangkou, Steward et al 541 (A, F, NY, S, US) ; Kweiyang, Tsiang 8653 (A, NY) ; 
Nganping, Handel-Mazzetti 132 (A) ; Tsingchen, Teng 90315 (A) ; Tsunyi, Steward et al 
97 (A, F, NY, S, US); Tuhshan, Tsiang 699 s (A, NY); Tungtze, Tsiang 4945, 4970 
(A, NY), shensi: Hua-Shan, Hers 3097 (A); Shih Tan, Hers 2442 (A); Tai-pei Shan, 
Purdom 946 (A, US) , Fenzel 7J<?(A) ;Tsi-yuan Hsien, Hers 2670(A) ;Tsinling Shan, Fenzel 
678 (A), Hers 2932, 3010 (A); Tze-wu Hsien, Meyer 1722 (A), szechuan: Li-fan 
Hsien, Wang 21670 (A) ; Kwang-yun Hsien and vicinity, Wang 22633 (A) ; inter Nitou 
et Hualinpin, Smith 1 3448 (UPS); Nan-chuan, Fang 571 (A), 1 362 (A, NY); Wen- 
chuan, Wilson 36^ (A) ; without precise locality, Bock & v. Rosthorn 790 (A), yunnan: 
Eche-hay, Maire 385 (A); Hsinlung, Handel-Mazzetti 5692 (A); Tong-tch-ouan, Maire 
22, 393 (A), 6224 (UC). 

This species has the largest leaves of any species of Celastrus. They reach 
as much as 16 cm. in length and 15 cm. in width. The characteristic leaves 
and angular branchlets with dense, small and whitish lenticels make it easily dis- 
tinguishable from allied species. I saw a living plant of this species cultivated at 
the Brooklyn Botanic Garden, New York, which has long and dense, paniculiform 
fruiting branches. During winter and early spring the dehiscing fruits with their 
crimson arillate seeds present a beautiful and eye-catching sight. This species would 
make a valuable ornamental plant although it is not now used extensively. 

7. Celastrus scandens L. Sp. PL 196. 1753; Gray, Gen. PL U. S. 2:186, 
/♦ 170. 1849, ex char. 

Celastrus bullatus L. Sp. PL 196. 1753, ex char. 
Euonymoides scandens (L.) Moench, Meth. PL 70. 1794. 



Scandent shrubs up to 20 m. tall; branches terete, glabrous, castaneous, the 
lenticels scattered, orbicular or elliptic, the current year's growth yellowish-green, 
with lenticels lacking or inconspicuous; axillary buds subglobose, about 1.5 mm. 
long. Leaves elliptic, ovate-elliptic to obovate, the apex acute to short-acuminate, 
the base rotund to cuneate, the margins crenate-serrulate, incurved, glandular- 
mucronate, 4-12 cm. long, 2-6 cm. wide, membranous, glabrous; the primary 
lateral veins 4-6 pairs, distinct and slightly elevated on both surfaces, the veinlets 
visible on both surfaces; stipules filiform, tufted, about 1.5 mm. long; petioles 
usually 1.0-2.5 cm. long. Inflorescences terminal, usually once compound, 2-8 
cm. long, the peduncles glabrous, the primary peduncles obscure to 5 mm. long; 
flowers dioecious, green, the pedicels obscure to 3.5 mm. long, the articulation at 
the upper half of the stalk. Male flowers: calyx lobes open, oblong, rotund, slightly 
erose, accrescent, about 1 mm. long; petals thin, more or less transparent, rotund, 
slightly erose, about 4 mm. long, 1.5 mm. wide; disc cup-shaped, the lobes trun- 
cate; stamens about 2.5 mm. long, the filaments slender, glabrous, the anthers ovoid, 
obtuse, apiculate; sterile pistil 1.2 mm. long. Female flowers: calyx lobes, petals, 
and disc as in the male; sterile stamens about 1 mm. long; pistil flask-shaped, about 
3 mm. long, the ovary ellipsoid, the style cylindric, the stigmata 3-lobed, patent. 
Fruiting inflorescences usually racemiform; fruits subglobose, the valves sub- 
orbicular or broadly elliptic, about 8-12 mm. long and 7-10 mm. wide; seeds 
ellipsoid or slightly plano-convex, obtuse at the base, attenuate toward the apex, 
about 4.5 mm. long and 2 mm. wide, brown, the areolae obscure. 

Roadsides and thickets; lowlands up to 600 m. alt.; North America; flowering 
from May to June. 

Canada: Ontario: Algoma, Bruce, Essex, Frontenac, Glengarry, Leeds, Renfrew, 
Waterloo, Welland, and York counties. Quebec: Brome, Chambly, Chateauguay, Mis- 
sisquoi, and Pontiac counties. 

United States: Arkansas: Benton County. Connecticut: Litchfield and New 
Haven counties. Delaware: Newcastle County. Illinois: Adams, Champaign, Fulton, 
Jackson, La Salle, Mason, Peoria, Schuyler, Stark, Vermi ion, and Winnebago 
Indiana: Floyd, Franklin, Gibson, Jennings, Jefferson, Kosciusko, Marion, Porter, Sullivan, 
Tipton, and Wells counties, iowa: Cedar, Clay, Davis Decatur Fayette Hardin, Lee, 
Story, and Winneshiek counties. Kansas: Cowley, Douglas, Ellis, Geary R.ley, and Wil- 
son counties. Kentucky: Boyd, Fayette, Greenup, Rockcastle, and Warren 
maine: Cumberland, Kennebec, Penobscot, and York counties. Maryland: Caroline, 
Cecil, Harford, Montgomery, and Talbot counties. Massachusetts: Berkshire, Dukes, 
Essex, Franklin, Hampden, Hampshire, Middlesex, Norfolk, and Suffolk counties. Michi- 
gan: Cheboygan, Emmet, Keweenaw, Menominee, Midland, and Oceana counties^ minne- 
sota: Clearwater, Goodhue, Hennepin, Houston, Kandiyohi Nicollet, and Otter Tail 
counties. Missouri: Atchison, Benton, Chariton, Clay, Cole, Franklin, . Hickory Howe I, 
Iron, Jackson, Jasper, Jefferson, Johnson Lafayette Lawrence, Oregon Ozark Polk Ra 11 , 

Warren, and Wright 



Nebraska: Antelope, Cass, Franklin, Lincoln, Sioux, Thomas, and York counties. 
Hampshire: Cheshire, Grafton, Hillsboro, Merrimack, and Rockingham counties 
jersey: Atlantic, Bergen, Burlington, Camden, Cape May Cumberland, Gloucester Hun- 
terdon, Mercer, Middlesex, Monmouth, Ocean, Salem, Somerset, Sussex, and Warren 
counties, new york: Albany, Cayuga, Columbia Dutchess, Orange, Saratoga, • Su^olk. 
St. Lawrence, Tompkins, Ulster, Warren, and Washington counties, north ««>una. 
Buncombe, Iredell, and Mitchell counties, north Dakota: Pembina, Ramsey, and Stark 

[Vol. 42 


counties, ohio: Clinton, Cuyahoga, Gallia, Hamilton, Harrison, Jackson, Lawrence, Meigs, 
Richland, Sciota, and Wood counties. Oklahoma: Blaine, Cleveland, Johnston, Kay, 
Murray, Payne, and Woodward counties. Pennsylvania: Adams, Allegheny, Beaver, 
Bedford, Berks, Blair, Bucks, Cambria, Carbon, Center, Chester, Clarion, Clearfield, 
Clinton, Crawford, Cumberland, Dauphin, Delaware, Erie, Fulton, Huntingdon, Indiana, 
Jefferson, Juniata, Lancaster, Lebanon, Lehigh, Luzerne, Lycoming, Monroe, Montgomery, 
Montour, Northampton, Northumberland, Perry, Philadelphia, Pike, Schuylkill, Snyder, 
Somerset, Sullivan, Susquehanna, Tioga, Union, Venango, Wyoming, Westmoreland, and 
York counties, rhode island: Providence County, south Dakota: Brookings, Brule, 
Grant, Lawrence, and Pennington counties. Tennessee: Crockett, Dyer, Gibson, Hamb- 
len, Obion, and White counties, texas: Armstrong, Brewster, Culberson, Jeff Davis, and 
Sutton counties. Virginia: Bedford, Caroline, Craig, Fauquier, Giles, Roanoke, Shenan- 
doah, Surry, and Isle of Wight counties. Vermont: Caledonia and Rutland counties. 
west Virginia: Cabell, Marshall, Summers, and Wyoming counties. Wisconsin: Brown, 
Buffalo, Door, Iowa, Kenosha, Lafayette, Outagamie, Polk, and Walworth counties. 
Wyoming: Crook County. 

This is the only widely distributed Celastrus species in North America. It is 
also commonly cultivated as an ornamental plant in the United States. This 
species is easily recognized by its paniculiform inflorescences, open calyx lobes, cup- 
shaped disc, and especially by the crimson arillate seeds of the opening fruits. 
Since it is widely distributed and the only commonly known species of Celastrus in 
North America, it will not be confused in the field or in the herbarium. I have 
seen many specimens of this species all of which are identified correctly. In the 
present treatment, I have not cited individual collections, but just listed the 
counties where they are found. All of the collectors and their field numbers can 
be found in the Index to Exsiccatae at the end of this paper. The variations of 
morphological characters and geographical distribution of this species have been 
discussed under the section on general morphology. 

Series 2. axillares Rehd. & Wils. in Sarg. PL Wils. 2:355. 1915. 

Sempervirentes Maxim, ex Rehd. & Wils. in Sarg. PL Wils. 2:357. 1915. 


A. Flowering branches terminal or rarely in some species those of the female plants 

B. Flowering branches terminal both in male and in female plants; inflorescences both 
lateral and terminal. 

C. Evergreen species, the current season's branches usually subtended by foliage 

leaves; inflorescences usually without accompanying vegetative buds (except 

some specimens of Celastrus hindsii); fruits 1 -seeded. 

D. Young branches with distinct lenticels; flowers numerous in dense umbelli- 

form cymes; discs subfleshy or fleshy, flat; leaves not densely reticulate. 

E. Fruit conspicuously stipitate; seeds about 15 mm. long, 8 mm. wide. 

India, Burma, Siam, Indo-China, southwestern and southeastern China. 

8. C. monospermus 

EE. Fruit not stipitate; seeds about 9 mm. long, 5 mm. wide. Malaya, Indo- 
nesia, Philippines and New Guinea 9. C. tnonospermoides 

DD. Young branches rarely with lenticels; flowers few in simple lax cymes, or 
solitary; discs membranous, cup-shaped; leaves densely reticulate; seeds about 

6 mm. long, 5 mm. wide. Southeastern Asia 10. C. hindsii 

CC. Deciduous species, the current season's branches not subtended by foliage leaves; 
inflorescences with accompanying vegetative buds; fruits 3- to 6-seeded. 
F. Flowers fasciculate or solitary, without distinct peduncles or with very short 
ones on the terminal inflorescences; calyx lobes imbricate; petals thin and 



with distinct venation, 2.5-3.0 mm. long, 1.0-1.5 mm. wide; anthers of the * 
male flowers 0.7-1.0 mm. long; fruiting inflorescences erect, the stalks of 
fruits not accrescent after anthesis, about 1.0—1.5 cm. long; seeds ellipsoid 
or ovoid. 

G. Terminal inflorescences usually 2-6 cm. long; axillary inflorescences usually 
several- to many-flowered, fruiting frequently. 
H. Veinlets of leaves usually immersed above and elevated beneath; seeds 

I. Anthers of male flowers not apiculate; disc lobes subquadrate to ob- 
long, obtuse or slightly emarginate; leaves usually paler beneath. 

Southwestern and central China and Indo-China 11. C. glaucophyllus 

II. Anthers of male flowers apiculate; disc lobes subreniform, mucronate; 

leaves concolorous. India, Burma, and China 12. C. hookeri 

HH. Veinlets of leaves elevated on both surfaces; seeds highly rugose. India. 

13. C. membranifolius 

GG. Terminal inflorescences usually 6-18 cm. long, axillary inflorescences few- 
flowered or suppressed, fruiting infrequently. Southwestern China 14. C. vanioti 

FF. Flowers in distinctly pedunculate dichasia on the lower two-thirds of the 
terminal inflorescences; calyx lobes valvate; petals relatively thicker and 
without distinct venation, about 4 mm. long and 2 mm. wide; anthers of the 
male flowers 1.5 mm. long; fruiting inflorescences pendulous, the stalks of 
the fruits accrescent after anthesis, about 2-4 cm. long; seeds slightly 

lunate. Central and western China 15. C. hypoleucus 

BB. Flowering branches terminal in male plants only; inflorescences both lateral and 
terminal in male plants, lateral only in female plants. 

J. Inflorescences dichasial, conspicuously pedunculate; pedicels of fruits green. 
K. Vegetative buds relatively large, ovoid, acuminate, 5-1 1 mm. long; leaves 
usually broadly ovate, rarely broadly elliptic, the secondary veinlets densely 
reticulate and elevated; anthers apiculate. China south of the Yangtze 
Rj ver 16. C. gemmatus 

KK. Vegetative buds relatively small, conoid, obtuse, 1-3 mm. long; leaves usually 
broadly elliptic, or broadly obovate to nearly orbicular, the secondary vein- 
lets slightly reticulate and elevated; anthers not apiculate. China north of 

the Yangtze River, Korea, and Japan - 17. C. orbiculatus 

JJ. Inflorescences usually fasciculate or subracemiform, usually sessile, rarely very 
shortly pedunculate; pedicels of fruits brown. 

L. Disc lobes broader than long, about one-fourth to one-fifth as long as the 
disc proper, the articulation usually at the middle or lower half of the stalk. 

China south of the Yangtze River - IS C rosthornianus 

LL. Disc lobes longer than broad, about half to one-third as long "the. dl * c 
proper, the articulation usually at the upper third or upper half of the 
stalk. Southeastern China, Formosa, southern Japan, and Ryukyu Islands 

19. C. punctatus 

AA. Flowering branches usually distinctly intercalary. 

M. Flowers usually in pedunculate cymes, axillary or extra-axillary; leaf margins ser- 
rate, the petioles less than half as long as the blades. 
N. Seeds slightly lunate, attenuate at both ends; sepals usually erose or glandular; 
secondary veinlets of leaves conspicuous or elevated on both surfaces. 
O. Leaves usually broadly elliptic to orbicular; disc subfleshy, flat or slightly 

concave. # ., , 

P. Plants glabrous to sliehtly pubescent on youne flowering branchlcts and 

the leaves beneath; leaves usuallv suborbicular or orbicular: disc lobes 

inconspicuous. China: Kwangsi, Kwangtung, Hainan, and Formosa 

20. C. kusanos 

PP. Plants densely yellowish brown-pubescent, especially on both surfaces of 
the leaves; leaves usually broadly ovate or broadly obovate; disc lobes con- 

spicuous, arcuate. Burma, China, and Indo-China t' 2l 'j ^rsutus 

OO. Leaves usuallv elliptic or elliotic-oblong; disc membranous, cup-shaned. 

Eastern India,' Indonesia, and China 22 ' C * 5 '> /<wt " 

NN. Seeds semi-annular, obtuse or rounded at both ends: sepals entire; secondary 

veinlets of leaves obscure on both surfaces. Southeastern China 23. C. aculeatus 

MM. Flowers usually solitary and axillary, or rarely fasciculate; leaf margins finely 
ciliate-serrate, the petioles nearly half as long as the blades. Northeastern and 
eastern China, Korea, and Japan 24. C. flagellant 

[Vol. 42 



Fig. 5. Celastrus monospermus Roxb. 

8. Celastrus monospermus Roxb. Hort. Beng. 18. 1814, nom. mid.; Fl. Ind. 
ed. Carey & Wall 2:394. 1824, ed. 2, 1:625. 1832 (as monosperma). (T.: 

Desiloa s.n. NY!). 


Catha benthamii Gardn. & Champ, in Hook. Kew Jour. 1:310. 1849, ex char. 
Champion s. n.). 

Celastrus championii Benth. in Hook. loc. cit. 3:334. 1851, ex char, (based on Catha 

benthamii Gardn. & Champ.). 
Celastrus benthamii (Gard. & Champ.) Rehd. & Wils. in Sarg. Pi. Wils. 2:3 58. 1916. 
Celastrus annamensis Tardieu, in Fl. Gen. l'Indo-Chine, Suppl. 1:803, fig. q8, t. 7. 1948, 

char. (T.: Poilane 24300). 

Scandent shrubs up to 10 m. tall; branches terete, smooth, glabrous, reddish or 
dark brown, the lenticels ovate to orbicular, sparse to dense, sometimes obsolete; 
axillary buds conoid, about 2 mm. long. Leaves variable, elliptic to oblong or 
broadly ovate, the apex acute to obtuse, the base cuneate to obtuse, the margins 
serrate, 7-17 cm. long, 3-9 cm. wide, firmly membranous, glabrous on both sur- 
faces, the primary lateral veins 5-8 pairs, curved toward the apex, elevated below, 
immersed above, the veinlets distinct, slightly elevated below, immersed and obso- 
lete above; stipules laciniate, about 1 mm. long; petioles 0.5-1.5 cm. long. In- 



florescences axillary, simply dichotomous or paniculiform, few and laxly branched, 
up to 20 cm. long, sometimes longer than the subtending leaf, distinctly peduncu- 
late, the peduncles usually glabrous, rarely pubescent, the primary peduncles 0.5-3.0 
cm. long; flowers dioecious, small, greenish-yellow or white, the pedicels obsolete 
to 4 mm. long, the articulation toward the base of the stalk. Male flowers: calyx 
lobes imbricate, semi-orbicular, slightly glandular-cilia te, about 1 mm. long; petals 
shortly oblong, about 2.5 mm. long and 2 mm. wide; disc fleshy, annular, slightly 
lobed, the lobes entire, subquadrate, broader than long; stamens attached just under 
the margin of the disc, about 2.7 mm. long, the filaments filiform, the anthers 
slightly ovoid, pink-punctate, slightly apiculate; sterile pistil 1.3 mm. long, ovoid, 
slightly immersed in the disc. Female flowers: calyx lobes, petals, and disc as in 
the male, usually the petals smaller than in the male; sterile stamens about 1 mm. 
long; pistil about 3 mm. long, the ovary subglobose, narrowed into the style, the 
stigmata 3 -lobed, reflexed. Fruits cylindric, stipitate, 1 -seeded, the valves ovate- 
oblong, about 2 cm. long and 1 cm. wide; seeds cylindric, about 1.5 cm. long and 
0.8 cm. wide, smooth, pinkish brown. 

Chiefly in thickets, at altitudes from 246 to 1,880 m.; commonly distributed 
in India, Pakistan, Burma, Indo-China, and China; flowering from March to June. 

Burma: Pang Hoi Phi and Peng Sai, Rock 2237 (A, US); without precise locality, 

Prazer s.n. (CAL). 


F, NY), 44169 (A, NY, US) ; Kan-en Hsien, Lau 371 1 (A, S) ; Po-ting, Haw 73564 (A, 
S) ; Yaichow, How & Chun 70273 (A, NY, US), Liang 62583 (A, NY) ; without precise 
locality, Liang 63419 (A, L, NY, S, US), 65253 (A, NY, S, US), Wang 35222 (NY), 
35893, 36064 (A, NY), 36607 (NY), kwangsi: Bako Shan, western Poseh, Ching 7525 
(A, UC, US); Me-kon, southern Nanning, Ching 8433 (A, NY, UC, US); Ping Nam 
Hsien, Wang 40415 (A); Shap-man-taai Shan, Shang-sze Hsien, Tsang 22079 (A, S), 
24106, 24188, 24735 (A, MO, NY) ; Yeo-mar Shan, northern Hin-yen, Ching 7 J 20 (NY). 
kwangtung: Kowloon, Wang 3056 (NY); Lofou Shan, Ho 60086 (NY), Ko 50089 
(NY), Tsiang 1705 (A, NY, UC) ; Lung-tau Shan, near Iu, Kan-peng To et al 473 (UC) ; 
Sin-fung Hsien, Taam 720 (A); Sunyi, Ko 51463 (A); Tai-o, Chun 3107 (NY); Tapu 
Hsien, Tsang 21751 (A, NY, S) ; Wung-yuen Hsien, Lau 800 (A, NY, SING) , 2750 (A) ; 
Ying Tak, Tsang # Wong 268 1 (UC), Tsui 348 (MO, NY), eastern tibet & south- 
western china: without precise locality, Forrest 26407 (A), yunnan: Che-li, Wang 
75920, 78458, 79015, 79459> 79776 (A); Chen Kang, Yu 17495 (A); Fo-hai, Wang 
73579, 73590, 73825> 73^39, 74430, 74907, 77*02* (A); Jenn-y eh Hsien, Wang 80312, 
80781 (A); Lan-tsang Hsien, Wang 73415 (A); Lung-ling, Tsai 55679 (A); Lu-se, Tsai 


56917 (A) ; Mengtze Henry 10446 (A, NY), 10955 (A, MU), 1 1399 (A, MO, INT, 
between Mohei and Maokai, Rock 2900 (A, UC, US) ; Nan-chiao, Wang 75183 
between Palut and Nam-dip, on the trail from Raheng to Mesawt, Rock 1024 (US); 
Ping-pien Hsien, Tsai 55076, 55139, 60152, 60459, 61680, 62295 (A); Shweli-Salwin, 
Forrest 1 1874 (UC) ; Si-chour Hsien, Feng II 445, ^53°, I2 4&9 (A); Sunning, Hila, Yu 
16508 (A) ; Szemao, Henry 11972 (A, NY), 11972* (A, MO, NY), Rock 2770, 2823 (A, 
UC,US) ; Tsing-pian, Tsai 52477,52552 (A) ; Tung-ting, Feng 13472,13749 (A) ; Wen-shan 
Hsien, Tsai 51546, 51591, 72857 (A); without precise locality, Wang & Ying Liu 82917, 
82947 (A). Hongkong: Chun 40143, 40276 (NY); Chun 4831 (VC),4943 (A, UC) ; 
Horn Fung 161 (NY); Taam 1 792 (A, US), 2087 (US); Lanto, Tsiang 720 (UC) ; 

Wright 9 (L), 92 (GH). _ . 

India: Assam: Darrang, Upendranath Kanjilal 3801 (CAL); Jawai, Dr. Kmgs col- 
lector s.n, (L); Khasi Hills, Kurz 182 (CAL); Kohima, Dr. King's collector 178 (U, 

[Vol. 42 


UPS); Konoma, Dr. King's collector 233 (A); Namchung, Luckimpore, Clarke 37949 s 
(CAL); Naga Hills, Jagarmani 499 (A, US); Dr. King's collector 795 (A, CAL), 1082 
(CAL); Silhet (as Sillet), Desiloa s.n. (NY, type of Celastrus monospermus) ; Hooker & 
Thomson s. n. (CAL, GH, L, S) ; Khasia, Liam, Clarke 451 52* (CAL); without precise 
locality, Griffith 666 (CAL). bengal: Sinchula, Biswas 2033 (A, NY); Darjeeling, 
Lace 2229 (CAL) ; Darjeeling, Anderson 107 (CAL) ; Wangiet, Kurz s. n. (CAL) ; Runquo 
Valley, Anderson 108, 109 (CAL); Sureil, Darsteri 17 (CAL). 

Indo-China: Chapa, Tonkin, Petelot 3184 (UC), 5931 (A), 6368 (A, NY, US). 

Pakistan: Griffith 1992 (CAL, GH, L, NY, S). 

This species is distinguished readily by stipitate fruit bases, the one-seeded 
capsules, the large cylindric seeds, the characteristic fleshy annular disc, and the 
attachment of the stamens immediately beneath the margin of the disc. 

The synonymy of this species is due to the failure, first, of Bentham to appre- 



variability of that species. My present interpretation is that of Lawson (in Hook. 
Fl. Brit. Ind. 1:618. 1875). 

Generally speaking, plants of Yunnan, Burma, and India have oblong or elliptic- 
oblong leaves, while those of Kwangtung and Kwangsi have broadly elliptic or 
broadly ovate leaves. At first I attempted either to maintain these plants as two 
species or to unite them into one species with two subspecies having differing leaf 
characters. Since then, I have measured the length and width and angles A (apical) 
and B (basal) 30 from the different areas, and found that the frequency curves of 
the ratio of length and width, and the angles A and B overlap. Since leaf shape is 
an extremely variable character and the floral structures of these plants are so 
similar, I feel justified in considering them as representing a single species. 

9. Celastrus monospermoides Loes. in Lorentz, Nova Guinea 8:280. 1910. 
(T.: Versteegh 1643, L!). 

7 onti 

No. 8:353). 1896, as to spec, cited, non Benth. 
Celastrus apoensis Elmer, Leafl. Philipp. Bot. 7:2579. 1915. (T.: Elmer 11411, A!). 
Celastrus malayensis Ridl. in Jour. Roy. Asiat. Soc, Str. Branch 75:18. 1917. (T.: Ridley 

13538, SING ! ) . 

Scandent shrubs up to 17 m. tall; branches terete, smooth to striate, glabrous, 
black or reddish-brown; lenticels lacking or obscure; axillary buds conoid, about 
2 mm. long. Leaves ovate to elliptic, the apex acute to acuminate, the base shortly 
cuneate, obtuse to rounded, the margins serrate to nearly entire, 5-16 cm. long, 
2.5-7.5 cm. wide, firmly membranous, lucid and usually blackish-brown on both 
sides or yellowish- green below in dry condition, the primary lateral veins 4-7 
pairs, curved toward the apex, conspicuous and elevated below, plane above, the 
veinlets distinct and slightly elevated below, immersed and obsolete above; stipules 
laciniate, about 1 mm. long; petioles 6-13 mm. long. Inflorescences axillary, soli- 

50 Woodson, R. E., Jr. Some dynamics of leaf variation in Asclepias tuberosa. Ann. Mo. Bot. 
Gard. 34:353-432. 1947. 




Fig. 6. Celastrus monospermoides Loes. 


tary or in clusters of three, aggregate dichasia, frequently fasciculate-racemiform, 
once to thrice compound, up to 10 cm. long, the peduncles glabrous, the primary 
peduncles obscure to 16 mm. long; flowers dioecious, white, the pedicels about 4 

>art of the stalk. Male flowers: calyx lobes 
imbricate, suborbicular, entire, about 0.7 mm. long; petals oblong, the apex 
rounded, about 2 mm. long and 1 mm. wide, entire, pink-punctate; disc fleshy, 
plane, the lobes obscure, truncate; stamens attached slightly under the margin of 
the disc, about 2.5 mm. long, the filaments filiform, glabrous, the anthers ovoid 
and obtuse; sterile pistil conoid, 2 mm. long. Female flowers: calyx lobes, petals, 
and disc as in the male, sterile stamens about 0.7 mm. long; pistil flask-shaped, 1.2 
mm. long, the ovary globose, the style slender, distinct, the stigmata discoid. Fruits 

the valves ovate or obovate to broadly 


12 mm. 

angular-ovoid, obovoid to subglobose, < 
elliptic, about 12-17 mm. long and 7-1 1 
long and 5—10 mm. wide, blackish-brown, smooth. 

In dense woods, at altitudes 100-3,300 m,; Borneo, Indonesia, Malaya, New 
Guinea, and Philippines; flowering from January to September. 

Borneo: western Koetai, Endert 3846 (L). British north Borneo: Kini Taki River, 
/. tf M. S. Clemens 31827 (A, L), 32052 (L); Marai Parai, /. £* M. S. Clemens 32860 
(L), 33182 (L, UC) ; Penibukan, /. tf M. S. Clemens 32120 (L), 50253 (L, UC) ; Upper 
Kinabalu, /. # Af. S. Clemens 27870, 20010 (L), 50641 (A), 51295 (A, L, UC). 

JL. 42 


Sumatra: Gaju & Alas Land, van Steenis 8717 (L), 96 jl (A, L); old jungle near 
the Ack Kanopan, Loendoet concession, Koeale, Bartlett 6864 (F, MICH, NY, US) ; 
Laboebom Batoe, Kota Pinang, Toroes 4009 (A, NY, UC, US). 

Malaya: selangor: Ridley 8228 (SING), negri sembilan: Sutu, Alims 1 605, 
2103 (SING), penang: Government Hill, Curtis 404 (SING), Ridley 404 (SING, para- 
type of Celastrus malayensis) ; Telom, Ridley 1 3538 (SING, paratype of C. malayensis) ; 
top of Penang Hill, Ridley 7127 (SING), perak: Batu Petch, Wray TO31 (SING, para- 
type of C. malayensis); Hijan, Haneeff & Jan 2466 (SING); Larut, Dr. King's collector 
5690 (L, SING, UC, UPS), 6928 (L, US); Maxwell's Hill, Curtis 2005 (SING), Ridley 
s.n., 1890 (A); Tapa, Wray 175 (L). Singapore: Sungei Loyang, Mat s.n., 1894 
( SING ) . 

New Guinea: dutch new guinea: Bali, Stresemann s.n., May-Aug. 1911 (L) : 
s. w. of Bernhard Camp, Idenburg River, Brass 1 2990 (A, L) ; Dalman, 45 km. inward of 
Nabire, Kanehira & Hatnsima T2J02 (A) ; southern New Guinea, via Resi Mts. to HeJlwig 
Mts., Versteegh 1643 (L, type of Celastrus monospermoides) . northeast new guinea: 
Ogeramnang, Morobe, J. # M. S. Clemens 4581, 5084, 6253 (A), papua: Boridi, Cart 
13036 (A). 

Philippines: Mindanao: Davao, Mt. Apo, Elmer 11411 (A, type of Celastrus apoensis 3 
GH, L, MO, NY, US); Mt. McKinley, Edano 900 (A, L, SING, UC) ; BukMnon, Mt. 
Dumalucpihan, Ramos & Edano 38961 (A, L, US); Mt. Katanglad, Sulit 3188, 3421, 
9917, 10086 (L). 

The wide distribution, within geographically isolated areas, and the morpho- 
logical variations of this species are the causes of several names. Specimens col- 
lected in the Malay Peninsula from lowlands up to 1,330 m. altitude have leaf forms 
varying from ovate to elliptic. They have thick and large leaves and long in- 
florescences. A specimen {van Steenis 9651, A, L) collected at altitude 3,000 m. 
from northern Sumatra has small and thin leaves and short inflorescences — about 
1 cm. long. 

The type specimen of Celastrus apoensis Elmer from Mt. Apo (alt. 2,000 m.), 
Mindanao, Philippines, has unusually small leaves. In addition, two other speci- 
mens (Sulit 9917 and 10086, L) collected at altitudes 1,700-1,800 m. from Mt. 
Katanglad, Mindanao, have larger leaves. The shapes and textures of leaves and 
inflorescences are similar to those from Malaya, Borneo, and New Guinea in similar 

There are several collections from Mt. Kinabalu, North Borneo, at altitude 
about 3,000 m. The leaf form is ovate with rounded base which is very distinctive 
and might be thought as a form of this species. In addition, there is a specimen 
(Endert 3846, L) collected at altitude 1,200 m. from east-central Borneo having 
elliptic leaves, acute at both ends. I think it may be possible, in addition to those 
mentioned above, to find more specimens with various leaf forms from different 

The leaves of the New Guinea specimens are usually elliptic, acute at both ends. 
They are quite similar to those from other localities in similar environments. I 
have made some studies and measurements of leaf characters. Except those from 
Mt. Kinabalu, Borneo, they can hardly be separated into groups. The shapes and 
sizes of fruits are variable also. As far as the shapes of fruits are concerned, those 
of the Malayan specimens seem to be broadly elliptic while those of the Philippine 
specimens are obovate. However, they display intermediate forms as well. 




Map 2. Distribution of seven species of Celastrus, Subgenus celastrus, Series II. 


10. Celastrus hindsii Benth. in Hook. Jour. Bot. 3:334. 1851, ex char. (T.: 
Hinds s. n. ) . 

Catha monosperma (Roxb.) Benth. in Hook. Lond. Jour. Bot. 1:483. 1842, quoad descr. 

& spec, cit., non sensu Roxb. 
Celastrus venulosus Wall. Cat. no. 4321. 1831, nom. nud. 
Celastrus racemulosus Hassk. Hort. Bogor. Descr. 1:1915. 1858, ex cnar. 
Flueggea serrata Miq. Fl. Ind. Bat. 1:356. 1859. (T.: Junghuhn s.n., L!). 
Celastrus monosperma sensu Benth. Fl. Hongk. 63. 1861, non Roxb. 
Celastrus cantonensis Hance, in Jour. Bot. 23:323. 1885. (T.: Hance 22I 9 I, A, a photo 

and a fragment of a leaf!). , 4A1 . tn-467 

Ofcsrras Wwi Benth. var. W>i Loes. in Engl. Bot. Jahrb. 29:444. 1901, 30.467. 

1902. (T.: Henry 3495, A!). 

[Vol. 42 



Celastrus oblongifolia Hay. Icon. PL Formos. 3:58. 1913. (T.: Hayata s.n., A!). 
Celastrus marianensis Koidz. in Bot. Mag. Tokyo 30:400. 1915, ex char. (T.: Koidzumi 

Celastrus axillaris Ridley, in Jour. Roy. Asiat. Soc. Malay. Branch, 1:56. 1923, ex char. 
Celastrus approximata Craib, in Kew Bull. Misc. Inf. 1926:349. 1926. (T.: Kerr 9941, 


Celastrus merrillii Tardieu, in Bull. Soc. Bot. France 95:180. 1948; in FL Gen. l'lndo- 
Chine Suppl. 1:805. fig. 98, t. 1,2 & J. 1948. (T.: Poilane 28/14, P!). 

Fig. 7. Celastrus hindsii Benth 

Scandent shrubs up to 19 m. tall; branches terete, glabrous, brown, rarely 
lenticelled, the branchlets smooth and green, the lenticels usually lacking on the 
current season's growth; axillary buds deltoid, about 1 mm. long. Leaves elliptic- 
oblong, obovate-oblong, elliptic or rarely broadly elliptic, the apex abruptly acute 
to acute, the base cuneate to obtuse, the margins remotely and obscurely serrate to 
serrate, 4.0-14.5 cm. long, 1.5-6.0 cm. wide, coriaceous, shining, glabrous, the 
primary lateral veins 6-9 pairs, the veins and veinlets usually distinctly elevated 
and densely reticulated on both surfaces, rarely slightly elevated or obsolete above; 
stipules laciniate, about 1 mm. long; petioles about 0.5-1.0 cm. long. Inflorescences 
usually axillary as well as terminal, sometimes axillary only, solitary, usually 3- to 
5 -flowered, or sometimes racemiform and longer than the subtending leaf, the 



peduncles glabrous, the primary peduncles almost obsolete to 5 mm. long, rarely 
up to 25 mm. long; flowers dioecious, white to pale yellow, the pedicels almost 
obsolete to 5 mm. long, the articulation on the upper half of the stalk. Male 
flowers: calyx lobes semi-orbiculate, imbricate, obtuse, ciliate, about 1.5 mm. long; 
petals oblong to obovate, obtuse, glandular-ciliate, 2-5 mm. long and 1.3-2.5 mm. 
wide; disc cup-shaped, the lobes usually deltoid, rarely slightly oblong, shortly 
acute, rounded, or slightly dentate; stamens arising from the margin of the disc 
proper, about 2.3 mm. long, the filaments subulate, glabrous, the anthers ovoid, 
obtuse, cordate; sterile pistil ovoid, about 1.5 mm. long. Female flowers: calyx 
lobes, petals, and disc as in the male, but the petals smaller; sterile stamens about 
1.5 mm. long, the ovary subglobose, the style columnar, the stigmata 3-lobed. Fruits 
ovoid to subglobose, the valves broadly ovate to suborbicular, about 8-11 mm. 
long and 7-9 mm. wide, usually 1 -seeded; seeds cylindric, 5-8 mm. long and 5 mm. 

wide, reddish-pink, smooth. 

Chiefly in thickets, at altitudes from 232-1,800 m.; Bonin Islands, Burma, 
China, India, Indonesia, Indo-China, and Siam; flowering from January to October. 

British North Borneo: Penibukan, /. 6 M. S. Clemens 30493 (A, L, UC). 

— _„ — — ^ ^ _ — J — 

Lap., Toppin 6154 (CAL). 
China: fukien: Amoy 




5601 (A); Changchow, Chung 1174, 1206 (A, UC) ; Chuanchow, Chung 1084 (UC) ; 
Foochow and vicinity, Sin Ging Tang 7039 (A), 7 17 5 (MICH); Kuliang and vicinity, 
Sin Ging Tang 6991 (MO, UC). Hainan: Bak Sa, Lau 26227 (A); Ching-mai Hsien, 
Lei 129 (L, NY, SING, UC, US), 440 (NY, SING, UC, US); Hung-mo Shan 
and vicinity, Lai area, Tsang & Fung 241 (A, US) ; Nodoa, McClure s. n. (A, MO) ; 
Yaichow, How 70368 (A, F, NY) ; without precise locality, Levine s. ». (A, F, 
MO, US), 484 (A, F, GH, MO), hupei: Chang-lo Hsien, Wilson 561 A (A); Enshih 
Hsien, Ho-Cb'ang Chow 1944 (A, NY); Ichang, Wilson 561 (A, F, GH, MO, 
US); Patung, Ho-Ch'ang Chow 548 (A), 705 (A, NY), Wilson 450 (A, NY); without 
precise locality, Henry 3241 (NY, type of Celastrus hindsii var. benryi), 3495 (A, GH, 
US), 3856, 349S A (GH), v. Kosthorn 9 (A), kwangtung: Au-tsai, McClure 3544 
(NY) ; Canton and vicinity, Hance 22191 (A, photo of type of Celastrus cantonensis, 
with a piece of leaf), Levine s.n. (C.C.C. nos. 180—A, UC; 44 I ~^ F '_ GH « MO _i US; 
1236— A, GH, MO, US), Merrill IOO70 (UC), Keng-p. 
wang, Kang-ping To et al 889 (US); Kochow, Tsiang 

Hsien, Tsang 21422 (A) ; Pok-Lor, Horn Fung A-511 (NY); Mn-tung risien, 1 aam 507, 
759 (A) ; Sun-wui Hsien, Tso tf Tsiang 2024 (A, F) ; Tai-o, Wang 3196 (NY) ; Tseng- 
shing, Tsang 20209 (NY); Tsung-hwa Hsien, Tsang 20604 (MO, NY), 25019 (A); 
Wung-yuen, Lau 693 (A, NY, SING), 2672 (A); Ying-tak Hsien, Tsui 391 (A, MO, 
NY, UC, US), Tsiang 1919 (SING, UC), Tsang tf Kam-chow Wong 2661 (UC) ; Wung- 
yuen, Lau 693 (A, SING), kweichow: without precise locality, Cavalerte 3976 (A). 
szechuan: Kiating Fu, Wilson 3324 (A); Omei Shan, Chiao & Far, 1 458, 8639,11712, 
l 49 8l y 15221 (A); Hu 7380 (A), Lee 4551 (A), Wilson 2307 (A, US) <, Wang 8096 (A). 
t AI wan: Aderu, Heito, Suzuki 11162 (TAI) ; Arisan, Faurie 1376 (A, TAI)^ Hayata s n 
(A, photo of type of Celastrus oblongifolius). yunnan: Che-h Hs.en, Wang 7584b, 
79467, 79470 (A) ; Chen-pien Hsien, Ko 56089 (A) ; Jenn-yeh Hsien, Wang 80136 80181, 
80232 (A); Lu-se, Tsai 56342 (A); Luh-shuen Hsien, Wang 8ll 7 7 (A); Mar-h-po, 
Tung-ting, Feng 13068, 13210, 13387, 13454, 134*4, 13*47 (A) ; Mengtze, Henry 10559 
(A, MO, NY, US), Tsai 5*375 (A); Si-chour Hsien, Feng Il6S 5> {^62 (A)- Szen^o, 
Henry 11972* (US); Tsing-pien, Tsai 5*545 (A), monokono: Chun 4840 ■ <A. UC); 
Hani 370 ! (GH) ; TW i 4 77, "47 (US) , /**, W (A, US) ; Tsang 77 (UC) ; Wight 
01 (GH). 611 (GH, NY). 


[Vol. 42 

India: assam: Darrang, Upendranath Kanjilal 4976 (CAL); Kegwima, Naga Hills, 
Clarke 41911° (CAL); Khasia, Tserra, Clarke 561 5 (CAL). sikkim: Duphla Hills, 

Lister 68 (CAL). 

Indo-China: Clio Gank, ? Helot 863 (A, UC), 991 (UC) ; Tonkin, ? Helot 340 
(NY, US). 

Indonesia: java: Thibeureum, Arsin 19568 (L); Tjibodas, Hallier 462 (A), Valeton 
102 (A), Koorders 42439P (L), Valeton s.n. (L, UC); Tjidadap, Tjibeber, Winckel 92?, 
98P (L), 1633? (L, UC), 1682P (L, SING); Mt. Waripin, Premyer, Forbes 961 (CAL), 
GH, L); without precise locality, Forbes 921 (CAL, GH, L), 980 (GH), 980* (CAL), 
Junghuhn 27, 10$ (L), s. n. (type of Flueggea serrata Miq., L), Korthals 747 (A). 
Moluccas: Amboina, de Vrise & Teijsmann s. n. (L). Sumatra: Kabajakan to Tretel, 
Walter & Bangham 870 (A, NY); Waldregion, Junghuhn s.n. (L). 

Japan: bonin islands: Hah-jima, Wilson 8300 (A). 

Siam: Kao Lem, Kerr 9941 (SING, type of Celastrus approximata) ; Mt. Doi Chang, 
Chiengma, Rock 1744 (A, UC, US). 

Bentham in 1842 determined a Hongkong plant, collected by Hinds, as Catha 
monosperma (Roxb.) Benth., based on Celastrus monosperma Roxb. Subsequently 
in 1851, he described the same specimen as representing a distinct species, Celastrus 
hindsH Benth. He commented in the discussion: "The form, size, and structure 
of the capsule are like those of C. [Celastrus] paniculata, without any narrowing of 
the base, as in C. monosperma and C. Championi, although it is always mono- 
spermous (by abortion) as in the two latter, not three-seeded as it usually is in 
C. paniculata. All these species belong to the true Celastri, not to Catha, to which 
I had erroneously referred them in my former paper". So he reduced Catha mono- 
sperma (Roxb.) Benth. (excl. syn. Roxb.) into synonymy under Celastrus hindsii 
Benth.; but in 1861 he again adopted Celastrus monospermus Roxb. for the Hong- 
kong plant and reduced his own Celastrus hindsii to synonymy. Actually, Celastrus 
hindsii Benth. is readily distinguished from Celastrus monospermus Roxb. in both 
vegetative and reproductive structures, as Bentham has pointed out in the discus- 
sion under Celastrus hindsii, and the two species should be maintained as distinct. 

The firmly membranous leaves usually are elliptic-oblong, shining, and densely 
reticulate; and the capsules usually one-seeded, which make this species very easily 
distinguished from the others. 

I have seen a photograph of the type specimen and fragment of a leaf from 
the type of Celastrus cantonensis Hance. This plant has leaves, venation, and 
inflorescences which match the description of Celastrus hindsii. 

The original description indicates that Celastrus hindsii var. henryi Loes. 
differs only from the typical variety by the larger and thicker leaves, and also the 
nearly immersed obsolete veins. Later (1916) Rehder and Wilson stated: "in 
addition to the characters given by Loesener the variety differs from the type 
species in having subsessile and very shortly peduncled cymes". In all the speci- 
mens seen, I have not been able to distinguish with certainty var. henryi from the 
typical variety. Also, this variety does not differ at all from the typical form of 
the species in geographical range. Thus, this variety is reduced to synonymy here. 

A specimen collected from Yunnan by Wang (no. 76880) has leaves like 
Celastrus hindsii whereas the fruits are similar to C. monospermus. This may be a 
hybrid between the two, if indeed they are distinct. 





Fig. 8. Celastrus glaucophyllus Rehd. & Wils. 

jcoPHYLLus Rehd. & Wils. in Sare. PL Wils 



Celastrus rugosus Rehd. & Wils. loc. cit. 2:349. 1915. (T.: Wilson iw6, A!). 
Celastrus orbiculatus sensu Tardieu, in Fl. Gen. Plndo-Chine, Suppl. 1:806. 1948, non 
Thunb., as to spec, cited. 

Scandent shrubs up to 2-5 m. tall; branches terete, glabrous, castaneous, tl 
lenticels scattered, elliptic; branchlets glabrous, yellowish-brown, the lenticels 

w about 


cm. wide, firmly membranous, glabrous, or pubescent on the veins below, rarely 
bullate, usually glaucous or pale beneath, the primary lateral veins 5-7 pairs, ele- 
vated below, immersed to slightly elevated above, the veinlets distinct, slightly ele- 
vated below, immersed and obscure above; stipules laciniate, filiform, about 1.5 mm. 

long; petioles 5 


long. Inflorescences axill 

llary inflorescences 3- to 7-flowered, the terminal ones racemiform, usua 

[Vol. 42 


cm. long, the peduncles suppressed or up to 2 mm. long, glabrous, each associated 
with an axillary bud except the uppermost one of the terminal inflorescences; 
flowers dioecious, green, the pedicels about 2 mm. long, the articulation at the 
middle or upper half of the stalk. Male flowers: calyx lobes valvate, ovate, rotund, 
glandular- ciliate, about 1.5 mm. long; petals oblong, rotund, slightly erose, 3-4 
mm. long and 1.0-1.5 mm. wide; disc membranous, the lobes ovate, obtuse; stamens 
2.5-4.5 mm. long, the filaments filiform, the anthers ellipsoid; sterile pistil 1 mm. 
long. Female flowers: calyx lobes, petals, and disc as in the male; pistil flask- 
shaped, 5 mm. long, the ovary globose, the style distinct, slender, the stigmata 3- 
lobed, the lobes patent. Fruits subglobose, the pedicels accrescent, about 3-5 mm. 
long, the valves suborbicular, 7-10 mm. long and 8-9 mm. wide, 3- to 6-seeded. 
Seeds ellipsoid to slightly plano-convex, about 3-4 mm. long and 2 mm. wide, 
black, the areolae distinct. 

Chiefly in mixed thickets, at altitudes from 700 to 3,300 m.; China and Indo- 
China; flowering from March to June. 

China: anhwei: Chu-hwa Shan, Ching 2700 (UC), Sun 1 189 (NY), hupei: 
Chienshih Hsien, Chow 1436 (A); Fang Hsien, Wilson 357 (A, MO, US), 972 (A). 
sikang: Kangting (Tachien-lu), Smith 1 3326 (UPS), Wilson 4117, 4122 (A); I-tung, 
Chiao 1721 (A); Tien-chuan Hsien, Tai et al 5142 (A), szechuan: Han-yuan Hsien, 
Fang 3737 (A) ; inter Hohsi et Telipu, Schneider 1 1 31 (A) ; Juei-she Hsien, Yu 941, 951 
(A); west of Kuan Hsien, Wilson 4317 (A); Mo-tien-ling, Wang 22462 (A); Mupin, 
Wilson 952 (A, holotype of C. glaucophyllus; GH, MO, US), 1148, 2310 (A, US), 3325 
(A); Ning-yuan-fu, Schneider 1027 (GH) ; Omei Shan, Chiao & Fan 767 (A), Chow 

6448, 6468 ', 7597, 8028 (A), Fang 3089 (A, NY), 12883, 13051 (A, US), 16624, '73*7 

(A), Hu 7380 (A), Lee 3561 (A), Tai 204, 1071 (A), Wilson 4782 (A) ; O-pien Hsien, 
Yu-shih Liu 2231 (A), Yu 750 (A) ; Ping-shan Hsien, Wang 22643 (A) ; Sungpan Hsien, 




us; MO) ; west of Wen 
Wa-shan. Wilson o<2* 

locality, Fang 1202 (A, NY), Schneider 646, 647 (A), yunnan: Che-tse-lo, Tsai 58555 
(A); Chien-chuan-Mekong Divide, Forrest 22235, 22475, 23230 (A, US); Chung- 
tien, Feng 3299 (A); western flank of Haba Snow Range, Feng 1226 (A); Hokin, 
Feng 787 (A) ; Kumming, Wang 62839 (A) ; Lan-ping Hsien, Tsai 54026 (A) ; Mengtze, 
Henry 9679 (NY, US); Pe-yen-tsin, Ten 40 (A), 550 (A, US); Ping-pien Hsien, Tsai 
61004 (A); south of Red River, Manmei, Henry 9679^ (A, MO, NY); Ta-pin-tze, Ten 
358 (A, US) ; Tengchuan, Schneider 2874 (A, GH, US) ; Likiang, Ching 20489, 20638, 
21618 (A) ,Feng 3047 (A) , Rock 3556 (A, UC, US) , 3931 (A, US) , 4032 (A, UC, US) , 
8312 (A), 8540 (A, NY, UC, US), Schneider 1984, 2839 (A, GH) ; Wei-se Hsien, Tsai 
57899>59824 (A). 

Indo-China: Chapa, Tonkin, Petelot 5941 (A, NY). 

12. Celastrus hookeri Prain, in Jour. Asiat. Soc. Bengal 73:179. 1904; Novic. 

Ind. 418. 1905. (T.: Hooker s. n., CAL!). 



Suppl. 1:806. 


Scandent shrubs up to 10 m. tall; branches terete to slightly striate, glabrous, 
brown to reddish-brown, the lenticels sparse or lacking on the current year's 
growth, suborbiculate to elliptic; axillary buds conoid, 2-4 mm. long, the bud 



scales accrescent, persistent. Leaves broadly elliptic to ovate, the apex acute, the 
base obtuse, the margins serrate, 6-12 cm. long, 4-7 cm. wide, thin-membranous 
on flowering branches, glabrous, the primary lateral veins 5-6 pairs, slightly ele- 
vated below, immersed and distinct above, the veinlets visible below, obscure above, 
relatively thick-membranous on fruiting branches; stipules laciniate, about 1.5 mm. 
long; petioles 8-15 mm. long. Inflorescences axillary as well as terminal, the axillary 
ones 3- to 5 -flowered, the primary peduncles glabrous, about 1.5-3.5 mm. long, 
the terminal inflorescences very short-pedunculate dichasia, fasciculate or solitary, 
about 4 cm. long, the axils of the peduncles associated with vegetative buds (except 
the uppermost) ; flowers dioecious, green, the pedicels 1.5-2.4 mm. long, the artic- 
ulation on the upper half of the stalk. Male flowers: calyx lobes deltoid, imbricate, 
cilia te, about 1.5 mm. long; petals oblong to obovate, slightly glandular on the 
margins, about 3.5 mm. long and 1.5 mm. wide; disc cup-shaped, the lobes sub- 
reniform; stamens arising from the margin of the disc proper, about 2 mm. long, 
the filaments linear, glabrous, the anthers apiculate; sterile pistil 1.5 mm. long. 
Female flowers: calyx lobes, petals, and disc as in the male, the sterile 
stamens 1.5 mm. long, the pistil 4 mm. long, the ovary globose, narrowed into a 
distinct style, the stigmata 3-lobed, each lobe bifid, linear. Fruits subglobose, the 
valves broadly elliptic, about 10 mm. long and 7 mm. wide, 3- to 6-seeded; seeds 
ellipsoid or ovoid, about 4 mm. long and 2 mm. wide, black, the areolae distinct. 

Chiefly in thickets, at altitudes from 1,525 to 3,050 m.; Burma, India, China, 
and Pakistan; flowering from March to May. 

Burma: Adung Valley, Kingdon Ward 9455 (A); Bhamo, Lapyeka to Sinlum Kaba, 
Lace 5775 (CAL); without precise locality, Prazer s. n., 1890 (CAL), Lace 5753 (CAL). 

China: yunnan: between Chienchuan Plain and the Mekong drainage basin to 
Lachining, Rock 8615 (A, UC, US); Wei-se Hsien, Tsai 59901 (A); without precise 

locality, Yu 8491 (A). f , ' 

India: assam: Khasia, Griffith 605 (CAL). bengal: Darjeeling, Clarke 6#2^ 9 
27046®, 35758* (CAL), Gamble 1926^ (CAL); Jalubaham, King s.n., May 9, 1876 
(CAL). sikkim: Lachung, Gammie 1179 (CAL) ; Lachen, Dr. King's collector s. n May 
1885 (CAL); Lebong, Kurz s.n. (CAL); Pauree, Dr. Prain's collector 289 (CAL); 
Lachen, Clarke 46569 (CAL). without precise locality: temp, region, 8,000-10,000 
ft., Hooker s. n. (type of Celastrus hookeri, CAL), Kurz s. n. (CAL). 
Pakistan: Griffith 1993 (CAL, GH). 

Rehder and Wilson 81 have cited two Indian specimens, Hooker d Thomson, 
"Khasia, alt. 4-5,000 ped." and C. G. Rogers, "Sikkim, January 1900", as Celastrus 

hookeri Prain. 

Wall. As it was stated by 


filaments in Celastrus stylosus, but glabrous ones in Hooker's species". In addition, 
the intercalary inflorescences and slightly lunate seeds of Celastrus stylosus are 


13. Celastrus membranifolius Prain, in Jour. Asiat. Soc. Bengal 73 : 197. 1904; 
Novic. Ind. 418. 1905. (T.: Mann s. n., CAL!). 
Branches terete, the branchlets slightly angular on the dry specimens, glabrous, 

:il Rehder, A. t and Wilson, E. H., in Sarg. PL Wils. 2:352. 1915 

[Vol. 42 


the lenticels scattered, ovate or orbiculate; axillary buds conoid, acute, about 1.5 


the margins serrate, 7.5-14.0 cm. long, 3.5-5.5 cm. wide, membranous, pale green 
in dry condition, glabrous, the primary veins 4-5 pairs, elevated below, immersed 
or slightly elevated above, the veinlets prominent on both surfaces; petioles 1.0-1.5 

cm. long. Flowers unseen. 


short-pedunculate or fasciculate, associated with axillary buds, the stalks 7-15 mm. 
long, the articulation on the upper half of the stalk. Fruits subglobose, the valves 
broadly elliptic, about 7 mm. long and 5 mm. wide, 3- to 6-seeded; seeds elliptic, 



India: assam: Khasia and Junteah Hills, Mann s.n., 1877 (CAL, type; MO); Khasia 
Hills and Brahmaputra plains, Kurz s. n. (CAL). 

The few specimens available are fruiting specimens in which the seeds are highly 
wrinkled, indicating that they might be immature. However, I have seen no 
immature seeds with such wrinkling in any other species of Celastrus. 

14. Celastrus vanioti (Levi.) Rehd. in Jour. Arn. Arb. 14:249. 1933. 

Saurauja vaniota Levi. Fl. Kouy-Tcheou, 415. 1915. (T.: Bodinier s. n., A!). 

Celastrus spiciformis Rehd. & Wils. in Sarg. PI. Wils. 2:348. 1915. (T.: Wilson 2U2, 

astrus spicif 

1915. (T.: Wilson u?6, A!). 

Scandent shrubs up to 10 m. tall; branches terete or slightly striate, glabrous, 
light brown to reddish or fuscous, the lenticels scattered, orbicular or elliptic, some- 
times lacking on the current year's growth, slightly elevated on the older branches; 
axillary buds globose, about 2 mm. long. Leaves ovate or elliptic, the apex acute 


serrate, the teeth incurved, glandular-mucronate, 5-13 cm. long, 3.5-7.5 cm. wide, 
membranous, glabrous or slightly puberulous on the veins below, the primary lateral 
veins 6-7 pairs, slightly elevated on both surfaces, the secondary lateral veins 
parallel; stipules laciniate, filiform, about 1.5 mm. long; petioles 1-2 cm. long. 
Inflorescences axillary as well as terminal, shortly pedunculate or fasciculate, the 
peduncles almost obsolete to 5 mm. long, the axillary inflorescences dichasia or very 




articulation at the middle or on the lower half of the stalk. Male flowers: calyx 
lobes deltoid, imbricate, obtuse, glandular-ciliate, about 1 mm. long; petals oblong, 
rotund, slightly erose, about 3 mm. long and 1.5 mm. wide; disc cup-shaped, 
membranous, the lobes acute; stamens arising from the margin of the disc proper, 
about 2.5 mm. long, the filaments linear, glabrous, the anthers ovoid, obtuse, 
slightly apiculate; sterile pistil conic, about 1 mm. long. Female flowers: calyx 



lobes, petals, and disc as in the male; sterile stamens 1 mm. long; pistil 3 mm. long, 
the ovary ellipsoid, gradually attenuate into the style, the stigmata 3-lobed and 
spreading. Fruits subglobose, the valves broadly ovate, about 9 mm. long and 8 
mm. wide, 3- to 6-seeded; seeds ellipsoid, obtuse at both ends, areolae distinct, 
about 4 mm. long and 2 mm. wide, black. 

In mixed woods, at altitudes from 500 to 2,000 m.; China; flowering from 

May to July. 

China: hunan: Wukang, Te-hui Wang 1 13 (A), hupei: Fang Hsien, Wilson 2215 
(A, NY); Heing-shan Hsien, Wilson 2312 (A, type of Celastrus spiciformis; GH, MO) 
without precise locality, Henry 5935 (A, GH, US), Wilson 2215* (NY), kweichow 
Bin Long, Miu Shan, Luchen, Ching 6057 (NY); Tuhshan, Tsiang 6655, 6993 (NY), 
without precise locality, Bodinier 2287 (A, type of Saurauja vanioti). szechuan: Loo- 
shan Hsien, Wang 23582 (A) ; Ma-pien Hsien, Wang 23WI (A) ; Omei Shan, Chow 7094 
(A), Tang 18958 (A), Sun tf Chang 149, 286 (A), Wang 8035 (A, US); Wa-shan, 
Wilson 1 176 (A, holotype of Celastrus spiciformis var. laevis), 1 38 3 (A, NY), yunnan: 
Menetze. Henry IIO06 (A, MO, NY): Tone Tch'ouan, Maire 6227 (UC); Wcn-shan 



somewhat intermediate in character between C. angnlata Maximowicz and C. 
hypoleuca Warburg". It appears to be a hybrid between those two species. Super- 
ficially, the type of C. spiciformis is similar to C. angulatus, except for the in- 
florescences which are both terminal and axillary. On examination, the floral 
structures were found to be similar to those of Celastrus hypoleucus. 

15. Celastrus hypoleucus (Oliv.) Warb. apud Loes. in Engl. Bot. Jahrb. 
29:445. 1900 (as hypoleuca). 

Erythrospermum hypoleucum Oliv. in Hook. Icon. Ill, 9:/. 1899. 1889, ex char. & ill. 

(T.: Henry 5877). 
Celastrus hypoglaucus Hemsl. in Ann. Bot. 9:150. 1895. (T.: Henry 2837, GH!). 
Celastrus hypoleucus forma a genuina Loes. in Diels, in Engl. Bot. Jahrb. 29:445. 1900 

(based on Erythrospermum hypoleucum Oliv.). 
Celastrus hypoleucus forma P argutior Loes. loc. cit. 1900. (T.: Henry 6771, A!). 

Scandent shrubs 3-5 m. tall; branches slightly striate, glabrous, darkish brown, 
the lenticels sparse to dense, rarely lacking on the current year's growth; axillary 
buds orbicular, about 2 mm. in diameter. Leaves elliptic to elliptic-oblong, the 
apex acute, the base obtuse, the margins remotely serrate to serrulate, 3.0-9.5 cm. 
long, 2.0-5.5 cm. wide, usually delicately membranaceous, glabrous, sometimes 
puberulous on the veins below, glaucous below, pallid above, the primary lateral 
veins 5-6 pairs, slightly elevated below, immersed above, the veinlets prominent 
below, obscure above; stipules filiform, about 2 mm. long; petioles 1-2 cm. long. 
Inflorescences axillary as well as terminal, the terminal portion usually up to 1 
cm. long, usually short-pedunculate, the peduncles glabrous, accrescent, the 


the pedicels accrescent, about 2-8 mm. long, the articulation at the upper half of 
the stalk, after anthesis usually only the flowers of the terminal inflorescences 
maturing into fruits, and the stalk accrescent up to 4 cm. long, spreading and 


[Vol. 42 

pendulous, Male flowers: calyx lobes valvate, deltoid to oblong, subentire to 
slightly erose, about 1.5 mm. long; petals oblong or rarely elliptic, usually entire, 
rarely obscurely erose, about 4.3 mm. long and 2 mm. wide; disc cup-shaped, the 
lobes obscure, truncate to mucronate; stamens arising from the margin of the disc 
proper, about 4 mm. long, the filaments filiform, glabrous, about 3 mm. long, the 
anthers ovoid, slightly apiculate, about 1.5 mm. long; sterile pistil about 2 mm. 
long. Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens 
about 1.5 mm. long; pistil flask-shaped, about 3 mm. long, the ovary ellipsoid, the 
style cylindric, distinct, the stigmata 3-lobed, flat, spreading. Fruits subglobose, 
the valves broadly elliptic or suborbicular, about 8-12 mm. long and 8—10 mm. 
wide, irregularly pink-punctate within, 3- to 6-seeded; seeds more or less plano- 
convex to slightly lunate, attenuate at both ends, about 6 mm. long and 2 mm. 
wide, black-brown, the areolae distinct. 

Chiefly in thickets, at altitudes from 1,700 to 2,745 m.; China; flowering 
from June to July. 

China: anhvei: Wang Shan, Ching 3065 (UC). hupei: Fang Hsien, Wilson 362 
(A, F, GH, MO, US); Hsun-tien-tsze, Chun 4427 (A); Ichang, Wilson 1063 (A, NY); 

without precise locality, Henry 2837 (GH, paratype of Celastrus hypoglaucus) , 5887 
(GH, NY, US), 68 1 1 (GH), 6771 (A, type of Celastrus hypoleucus forma P argutior; 

US), shensi: Kin-qua Shan, Giraldi s. n., July 10, 1897 (A) ; Qua-in Shan, Giraldi s. n., 

July 16, 1897 (A); Tai-pei Shan, Giraldi s. n., July 15-20, 1897 (A); Tsinling, Fenzel 

927 (A), szechuan: Kuan Hsien, Wilson 2306 (A, US). 

This species is easily recognized by its long terminal pendulous racemiform 
dichasia of long accrescent stalked fruits and by the slightly lunate seeds. The 
leaves are usually glaucous below, but this character is extremely variable. 

16. Celastrus gemmatus Loes. in Engl. Bot. Jahrb. 30:468. 1901 (as gemmata) . 
(T.: Henry 9872^ MO!). 

Embelia esquirolii Levi, in Fedde, Rep. Sp. Nov. 10:374. 1912. (T.: Esquirol 4, A!). 
Celastrus lokchongensis Masamune, in Trans. Nat. Hist. Soc. Formosa 25:15. 193 5. (T.: 
Tsiang 1 3 46 ', A ! ) . 

Scandent shrubs, 3-7 m. tall; branches terete or slightly striate, light to darkish 
brown, the lenticels scattered, suborbicular or oval, usually elevated, white; 
axillary buds conoid, acuminate, about 4-11 mm. long. Leaves usually broadly 
elliptic to elliptic-ovate, the apex gradually acuminate to acute, the base obtuse, 
rotund or truncate, the margins serrate, 5-15 cm. long, 2-8 cm. wide, firmly 
membranous, usually glabrous, rarely yellowish-puberulous on the veins below, the 
primary lateral veins 5-7 pairs, curved toward the apex, the veinlets densely retic- 
ulate, the veins and veinlets elevated on both surfaces; stipules filiform, tufted, 
about 1.5 mm. long; petioles 1.0-2.5 cm. long. Inflorescences axillary, usually 
3- to 7-flowered, the peduncles glabrous, the primary peduncles 4-7 mm. long; 
flowers dioecious, white or yellowish-green, the pedicels 3-8 mm. long, the articu- 
lation at the lower half of the stalk. Male flowers: calyx lobes valvate, ovate- 
deltoid, obtuse, glandular-ciliate, about 1 mm. long; petals subequal, obtuse, 




Fig. 9. Celastrus gemma tus Loes 

slightly erose, about 3.5-4.5 mm. long and 1.5-2.0 mm. wide; stamens inserted 
between the disc lobes, about 3 mm. long, the filaments more or less complanate, 
glabrous, gradually tapered toward the apex, the anthers oblong-ellipsoid, obtuse, 
usually apiculate, pink-punctate; disc thin, cup-shaped, the lobes acute; sterile 
pistil columnar, 1.5-2.5 mm. long. Female flowers: calyx lobes and petals as in 
the male; sterile stamens about 1 mm. long; disc thin, cup-shaped, the lobes in- 
conspicuous; pistil flask-shaped, about 4 mm. long, the ovary ovoid, the style 
columnar, the stigmata 3-lobed, each bifid, reflexed. Fruits usually 1-3 on each 
dichasium, the pedicels green, the fruits subglobose, the valves broadly oval or 
suborbicular, 7-15 mm. long and 7-15 mm, wide, 3- to 6-seeded; seeds ellipsoid 
or ovoid, about 4.5 mm. long and 2.5 mm. wide, reddish-brown, smooth. 

Common in thickets, valleys or open slopes, at altitudes from 400 to 3,000 m.; 
widely distributed in China: south of the Yangtze River; flowering from April to 


China: anhwei: Chu-hwa Shan, Cbing 2638 (A, UC), 7566 (UC); Wang Shan, 
Ling 1159, 1246 (UC); Wu-yuen, Ling 1 349 (UC). chekiang: east Tien-mu, Hu 1 585 
(A, UC). fukien: Dingschou, Tienhwa-schan, Wang 399 (A), hupei: Fang Hsien, 
Wilson 919 (A, NY), 2215 s - (A) ; Sin-shan Hsien, Wilson 363 (A, MO, US), 502 (A, F, 

[Vol. 42 


GH, MO, US); Patung Hsien, Ho-Ch'ang Chow 901 (A, NY); Wan-tsao Shan, Chun 

3884 (A), 4229 (US); without precise locality, Henry 761 4 (GH). kiangsi: Hsin-feng 

Hsien, Hu II 19 (A); Lu-shan, Chung & Sun 592 (A, NY), Steward 1234 (A), 5470 

(A, UC, US), Wilson 1 519 (A, MO, US), kiangsi et fukien: in monte Dunghwa-schan 

inter Schitscheng et Ninghwa, Wang 299 (A), kwangsi: Tunghua Shan, Thwang, Tsiang 

IOO47 (NY); Tzu-yuen Hsien, Chung 83605 (A); Huangtung, Yao-shan, Sin 11835 

(NY), kwangtung: Jen-hwa Hsien, Tsang 26404 (A); Lokchong, Chun 42959 (A), 

Tsang 20855 (A, L, MO, NY, SING, UC, US), Tsiang 1 346 (A, isotype of C. lokchong- 

ensis, UC), 1419 (A, UC) ; Tsungfa, Tsang 25189 (A); Yang-shan Hsien, Tsui 763 

(MO, NY), kweichow: Nan-kan, Cheng-feng, Tsiang 4661 (A, NY) ; Pinfa, Kweiting, 

Tsiang 5360 (A, NY); Tsingchen, Teng 9° 2 54 (A); Tungtze, Chun 5013 (A, S, UC), 

Tsiang 501 3 (NY); Yao-ren Shan, Sanhoa, Tsiang 6450 (A, NY); without precise 

locality, Esquirol 4 (A, type of Embelia Esquirolii). sikang: Kangting (Tachien-lu), 

Smith 12898 (S), Wilson 1302 (A, MO, US), szechuan: Kuan Hsien, Fang 2124 (A, 

NY); Mo-tien-ling, Wang 22451 (A); Monkong-ting, Wilson 2305^ (A); Mow Hsien, 

Fang 5529 (A); Nan-chuan Hsien, Fang 593 (A); Omei Shan, Chow 9906 (A), 

Fang tf Lee 3575 (A, US), Yu-shih Liu 1147, 1423 (A), Tai 1113, 1120 (A); O-pien 

Hsien, Sun 781 (A); Pao-Hsing Hsien, Chu 3761 (A); Wen-chuan Hsien, Wilson 2305 

(A, MO, US) ; Yun-ching Hsien, Sun 1247 (A), yunnan: Chien-chuan-Mekong Divide 

Forrest 21464, 22311 (A, US); Chungtien, Feng 2868 (A), Rock 24679 (A, MO, NY 

UC, US); in the mountain above Dashao, Handel-Mazzetti 13074 (A); between Kambaiti 

and Tengyueh, via Kuyung, Rock 7568 (A, UC, US) ; Kunming, Wang 62990 (A) 

Likiang Ching 20682 (A), Rock 3799 (A, UC, US), 4195 (A, US); Mengtze, Henry 

9782 (A, NY), 9782^ (A, type of C. gemmatus; MO), 10531 (A, paratype of C. 

gemmatus; MO, NY, US), 11471 (A, paratype of C. gemmatus; MO, NY) ; Muli Wachin, 

Yu 14375 (A); Ping-pien Hsien, Tsai 61004 (A); south of Red River from Manmei, 

Henry 9679^ (A); between Shweli and Yengyueh valleys, Forrest 8704 (A, S), II 173 f 

12050 (A) ; Suen-oui, Maire 14 (A) ; Tali, Rock 6809 (A, US) ; Tai-pon, Maire 102 (A) ; 

Wei-si Hsien, Tsai 59523, 59799, 63116 (A), Wang 63709 (A) ; without precise locality, 

Ducloux 132 (NY, UC), Forrest 16241, 14510 (A), Yu 5562, 5771 (A). 

Celastrus gemmatus often has been confused with C. orbiculatus Thunb., since 
the two are closely related. After a careful examination of specimens of the two 
species, it was found that although their floral morphology is similar, C. gemmatus 
has larger conic axillary buds, firmly membranaceous and densely reticulated 
leaves, and, in addition, a distinct geographical distribution which separates it 
from Celastrus orbiculatus and other related species. Celastrus gemmatus is chiefly 
distributed in southwestern China and is rarely found in southeastern China; some 
plants are recorded as far north as Hupei which may represent its northern limit. 

I have seen the types of Embelia esquirolii Levi, and C. lokchongensis Masamune, 
They both bear large conical axillary buds and densely reticulated leaves, and are 
within the geographical range of Celastrus gemmatus. Hence, I consider these 
species as synonyms of Celastrus gemmatus. 

17. Celastrus orbiculatus Thunb. Fl. Jap. pp. xlii, 97 (errore ft articulatus yy ). 
1784; Gmel. Syst. Veg. 406. 1796; Lam. Tab. Encycl. et Meth. Bot. 2:94. 
1797; Airy Shaw, in Curtis's Bot. Mag. l5S:tab. 9394. 1935. (T.: collector 
unknown s. n., A, photo! ) . 

Celastrus articulatus Thunb. Fl. Jap. 97. 1784 (sub. Celastrus orbiculatus, quoad nomen 
apud Thunb. in Fl. Jap. p. xlii); Maxim, in Bull. Acad. Sci. St.-Petersb., Ill, 27:456. 
1881; Mel. Biol. Acad. St.-Petersb. 11:200. 1881. 


18 57, ex char. 



Celastrus punctatus sensu Regel, in Gartenfl. 9:407, tab. 312, f. 6. I860, non Thunb. 
Celastrus articulatus Thunb. var. pubescens Makino, in Bot. Mag. Tokyo 7:102. 1893, ex 

char. (T.: Makino s.n.). 
Celastrus orbiculata Thunb. forma £. microphylla Loes. in Engl. Bot. Jahrb. 30:469. 

1901, ex char. (T.: Henry 3827). 
Celastrus orbiculata Thunb. forma 7. maior Loes. loc. cit. 30:469. 1901, ex char. (T.: 

Ciraldi 237) . 

Wils. in Sare. Pi. Wils. 2:350. 1915. 

( T. : Wilson 2308,Al). 
Celastrus lancifolia Nakai, in Bot. Mag. Tokyo 37:3. 1923, ex char. (T.: Nakai s. «.). 
Celastrus insularis Koidz. in Bot. Mag. Tokyo 39:22. 1925, ex char. (T.: Koidzumi s. ».). 
Celastrus strigillosus Nakai, loc. cit. 40:492. 1926, ex char. (T.: Nakai s.n.). 
Celastrus articulatus Thunb. var. stephanotiifolius Makino, in Jour. Jap. Bot. 3:24. 1926, 

ex char. (T.: Makino s.n.). 
Celastrus stephanotiifolius (Makino) Makino, loc. cit. 3:46. 1926. 
Celastrus orbiculata var. aureo-arillata Honda, in Bot. Mag. Tokyo 45:422. 1931, ex char. 

(T.: D. Shimisu 50) . 
Celastrus jeholensis Nakai, apud Nakai & Kitagawa, in Rept. 1st. Scientif. Exped. Man- 

choukuo, Sect. IV, 1:6, pi. I. 1934, ex char. (T.: Nakai et al. s. «.). 
Celastrus articulatus var. papillosus Nakai ex Hara, in Jour. Jap. Bot. 10:84. 1934, ex 

char. (T.: H. Hara s.n.). 

astrus articulata var. orbiculata (Thunb.) "Wang 


Celastrus orbiculatus var, papillosus (Nakai) Ohwi, Fl. Jap. 736. 1953. 

Celastrus versicolor Nakai, in Bull. Sci. Mus. Tokyo 33:16. 1953, ex char. (T.: Nakai 6 

Maruyama s.n.). 
Celastrus orbiculatus var. pilosus Nakai, loc. cit. 33:16. 1953, ex char. (T.: Hozawa- 

Sigeo s.n.). 
Celastrus orbiculatus f. papillosus Nakai ex Hara, Enum. Spermat. Japon. 80. 1954. 

Scandent shrubs up to 10 or 12 m. tall; branches terete to slightly striate, 
glabrous, light to darkish brown, the lenticels inconspicuous or sparse; axillary 
buds small, depressed, ovoid to subglobose, 1-3 mm. long, sometimes the outermost 
scales becoming deltoid, sharp-spinose, 1-2 mm. long. Leaves extremely variable 
in size and shape, usually obovate to suborbicular, ovate, or oval-oblong, the apex 
rounded, shortly cuspidate to acute, the base cuneate to obtuse, the margins crenate- 
serrate, 2-12 cm. long, 1.5-8.0 cm. wide, delicately membranaceous, glabrous 
above, rarely puberulent below, the primary lateral veins 3-6 pairs, distinctly 
elevated below, slightly elevated above, the veinlets slightly reticulated, prominent 
and slightly elevated below, immersed and obscure above; stipules filiform, tufted, 
about 1-2 mm. long; petioles 1-3 cm. long. Inflorescences axillary, rarely also 
terminal in the male plant, usually 3- to 7-flowered, occasionally solitary, often 
rather congested along the shoots of the current year, the subtending leaf some- 
times not developed, distinctly pedunculate, the peduncles subequal, glabrous, the 
primary peduncles 3-8 mm. long; flowers yellowish-green, regularly dioecious, but 
occasionally monoecious in late growth, the pedicels about 2-3 mm. long, accres- 
cent, the articulation usually at the base or lower third of the stalk. Male flowers: 
calyx lobes open, ovate-deltoid, subacute to obtuse, glandular-ciliolate, about 1.5 
mm. long; petals narrowly oblong to oblanceolate, obtuse, subentire to slightly 
ciliolate or erose, usually 3-5 mm. long and 1-2 mm. wide; disc thin, cup-shaped, 
the lobes usually acute; stamens arising from the margin of the disc proper, about 


[Vol. 42 


3 mm. long, the filaments filiform, glabrous, the anthers oblong-ellipsoid, obtuse, 
pink-punctate; sterile pistil columnar, about 2 mm. long. Female flowers: calyx, 
petals, and disc as in the male, sometimes the petals smaller; sterile stamens 
about 1.3 mm. long; pistil flask-shaped, about 4 mm. long, the ovary ovoid, nar- 
rowed into a columnar style, the stigmata 3-lobed, the lobes flat, recurved, some- 
times each slightly bifid. Fruits subglobose, bright yellow, the valves broadly 
elliptic to suborbicular, about 6-8 mm. long and 5-7 mm. wide, 3- to 6-seeded, 
the pedicels usually green in dry condition. Seeds usually ovoid or ellipsoid, about 
5 mm. long and 3.5 mm. wide, minutely areolate, pinkish-brown. 

Common in lowland slopes or in thickets, at altitudes from 100 to 1,400 m.; 



China: chekiang: Hai-wei, southern I-shing, Ching & Tso 580 (A), harbin: 
(Pinkiang), along a river, Skvortzov s.n., Dec. 7, 1929 (A), honan: Siashih, Hers 345 

?J ; ™ n 8 hsien ' Shih-tze-miao, Hers 1272 (A) ; Teng-feng Hsien, Yu-tai Shan, Hers 2701 
(A); Tsi-yuan Hsien, Hers 1758, 2705 (A); Yungning, Yu-tze-ping, Hers 828 (A); 
without precise locality, Hers 26 (A), hopei (chili): Ch'o-K'ou-tien, 50 km. sw. 
Peking, Bohlin 104 (S, UPS); Chang-li, Dorsett tf Morse 7152 (US); western hills, 
Dorsett & Morse 7206 (UC, US); Prince Park, Wang 176 (NY); mont. a l'ouest Peking 
pagoda, Pi-yun-sen, Janet 1639 (UPS); Haiao-wu-tai Shan, Smith 740, 959 (S, UPS); 
Kuan-tso-hng, Liu 1176 (UC) ; Ming Tombs, Nankou, Chiao 2126 (US) ; Tsing-ling-chiao, 
Chtao 21 16 (NY), Dorsett gf Mow 7074 (US); near eastern Tombs, Liou 385 (NY); 
western Tombs, Liu 398 (UC) ; without precise locality, Li 11169 (NY), hupei: Ichang, 
Chun 3505 (A) , Wilson 2308 (A, holotype of Celastrus articulata var. cuneata Rehd. & 
Wils.); Noh-chen Hsien, Chow 369 (NY); Lo-kia Shan, Wuchang, Sun 108 (NY); 
western Hupei, without precise locality, Wilson 181 (GH, US), jehol (gehol): David 
W (GH, US), kangsu: Mien Shan, Lin-shih Hsien, Tang 889 (A); Tienschui, 
Fenzel 2827 (A), liaoning: Liaoyang, Mukden, Li 69 (UC). port Arthur: Wilson 
,, e ( tt2: N j VNKING: Lin S H02 (UC). shansi: Chioh-hsin Dist., Smith S745 

AS, UPS) , 7697 (A UPS); Chin-yuan, Ling 1607, 1664 (UC); Fu-ping, Tao-ho-tze, 
Janet A34S (UPS) ; Fu-ping Ta-wu-tai Shan, Janet A349 (UPS) ; Hweihsien, Shansi 

~™ e x r * 5 P " 7 °? ( ^. } ; Tsin S- hu /« /""' A 79 (UPS) ; Dao-hui-gou, Wenshui, Wang 1 74 
(GH); Yuan-chu Dist., Yang-shu-ling, Smith 6154 (UPS), shangtung: Meng Shan, 
Fei Hsien, Cheo ef Yen 193 (A); Tsing-tao, First Park, Chiao 2395 (A, NY, US), 2612 

(A 'x F ; ?\\y^^J Tsin S- tao ' Lioent *3262 (A) ; Tsinan, Ching-lung Shan, Chiao 
3010 (A, NY, UC, US), shensi: Lao-y-san, Giraldi s.n., June 5, 1897 (A, UC); Tai- 

pe,-shan, Purdom 944 (A, US), 945 (A); Tsingling-shan centr. inter Mei et Liupa, 
Fenzel 502 (A). 

Japan: Honshu: Chiba Pref. Mobara, Walker 5717 (US) ; Kanagawa, Hakone in 
Sagami, Ohwt ? Okamot 504, 506 (L, S, UC, US); Hakone. Sawada 2t6t 


UC, Ub); Hakone, Sawada 2161 (S); Mt. 
), Shiota 3899, 3905, 9643, 9766 (A) ; Prov. 
44 (US); Prov. Nagano. Kimura ?07 (US). 

s. «., 1914 (A) ; Otake-gawa, Shinano Prov., Wilson s. n., 1914 (A) ; Suruga Prov., Wilson 
0907 (A, MO, US) ; Oku-shiobara, Tochigi, Suzuki 444020 (UC) ; Tokyo, Faurie 6134 
(A), Yamazala 44-A (US); Kamohura, Sagami Prov., Wilson 6627 (A, US); KumashH, 
Toman , Toyama, Bergman 439 (S) ; Prov. Ugo, Mizushima 1929 (MO); Mitsutooge, 
Yamanashi Pref., Suzuki 486024 (UC); Mt. Hoowoo, Yamanashi, Suzuki 491027 (UC). 
hokkaido: Prov Iburi, Zezo, Takenauchi s.n., July 17, 1916 (S) ; Sapporo, Faurie 6133 
A), Jack s. n., Aug. 22, 1905 (A, GH), Miyabe s. «., June 1880 (A, GH), Wilson 74 J2 
(A) ; vicmity of Obihiro, Tokachi, Dorsett d Morse 1070 (A, US) ; southern Hokkaido, 



without precise locality, Brooks 44 (A, UC), 544 (UPS), kurile islands: Shikotan, 
Ohwi 436 (UPS), without precise locality: Thunberg s.n. (A, type photo of C. 

orbiculatus, pi. 961). 

Korea: Port Chusan, Keisho Nan, Wilford s.n., 1859 (GH, S); Port Hamilton, Wil- 
fords. «., 1859 (GH, S) ; Keijo (Seoul), Mills s. «., Oct. 2, 1915 (UC) ; Kongo-san, Prov. 
Kogen, Wilson 10461 (US); Pyongang, Wilson 9203 (US); Taiyudo Prov., Hsian Hoku, 
Wilson 861 1 (US) ; without precise locality, Gilbert 6 (UC). 

Thunberg describes this species (Fl. Jap. p. 97, 1784) as Celastrus articulatus. 
However, in his index, p. xlii, articulatus is omitted and orbiculatus is listed. In 
1796, Gmelin 32 used the name orbiculatus and referred to Thunberg's Fl. Jap. p. 97, 
but he did not mention articulatus. Later, in 1881, Maximowicz pointed out 33 
that the name articulatus was due to a printer's error, having been changed from 
C. orbiculatus, as shown in the 'Flora Japonica' and the figures of Thunberg's 
manuscript. However, Maximowicz chose Celastrus articulatus as the species 
name because it was universally accepted at that time. Airy Shaw 34 states in 1935 
that "The form articulatus is a 'vain tradition', and the sooner it is dropped the 
better— but tradition dies hard!" I agree that it is best to use the originally correct 
name for this species, even more so because "orbiculatus" refers to the leaf shape. 

This is a very widely distributed eastern Asiatic species. It is found in northern 




specimens that have been collected south of the river ( northern Chekiang ) . 


eastern United States where it was introduced at the New York Botanical Garden 
in 1891 from seeds secured from the Royal Botanic Gardens, Kew 35 . The fruit- 
clusters are axillary and remain attractive through the winter after the leaves have 
fallen; the fruiting branches are as showy as those of the paniculiform Celastrus 
scandens. Birds are fond of the ripe seeds 36 and help to propagate them. 

I have observed two living plants of this species cultivated in the Missouri 
Botanical Garden. In the spring both have male flowers; all the floriferous branches 
bear both axillary and terminal inflorescences. In June, the male flowers fall, and 
the terminal inflorescences fall away from the plant; some of the axillary buds or 


bear onlv female flowers. Some 

which in these plants are scarce. The leaves are variable. Due to the variation 
caused by these characteristics, there are many synonyms for this species. 

'"Syst. Veg. p. 406. 1796. 

"Mel. Biol. Acad. St. Petersb. 11:201. 1881. 

34 Curt. Bot. Mag. 158:/**. 9394- 1935. 

35 Nash, C. V. Celastrus articulatus. Addisonia 4:9-10, pi. J 25. 1919. 

86 Burbidge, F. W. Japanese tree-strangler, or staff tree (Celastrus arttculatus) . Card. Cnron. 

Ill, 23:28, /. //. 1898. 


[Vol. 42 


Map J. Distribution of ten species of Celastrus, Subgenus celastrus, Series II. 





1900. (T.: Ros thorn 1572, A, photo!). 


Celastrus stylosa sensu Levi. Fl. Kouy-Tcheou, p. 69. 1914, non Wall. 
Celastrus loeseneri Rehd. & Wils. in Sarg. PL Wilson. 2:3 50. 1915. (T.: Wilson 357*, A!). 
Celastrus cavaleriei Levi, in Monde des PL II, 18:31. 1916. (T.: Cavalerie 496, A, photo!), 
non Levi. 1914. 

Scandent shrubs up to 7 m. tall; branches terete, occasionally striate, glabrous, 
gray-brown to brown-red, densely to sparsely lenticelled, the lenticels elevated, 
ovate; axillary buds ovoid, about 3 mm. long. Leaves elliptic, oval, or obovate- 
oblong, the apex acute, the base cuneate to obtuse, the margins remotely serrulate 
to serrate, usually 4-11 cm. long, 2-6 cm. wide, thinly membranaceous on the 
flowering specimens, firmly membranaceous on the fruiting specimens, glabrous, 
the primary lateral veins 4—7 pairs, prominent, slightly elevated on both surfaces, 
the veinlets obscure on both surfaces; stipules filiform, tufted; petioles 0.5-1.5 cm. 
long. Inflorescences axillary as well as terminal in the male plants, axillary only 
in the female, the terminal portion sometimes up to 5 cm. long, usually fascicular, 
rarely solitary; peduncles glabrous, the primary peduncles very short to nearly 
obsolete in the fascicular inflorescences, 2—5 mm. long at the solitary ones; flowers 
dioecious, yellowish-green, the pedicels 2-5 mm. long, the articulation at the 

or lower half of the stalk. Male flowers: calyx lobes valvate, ovate to 



stamens arising from the margin of the disc proper, 2.5 mm. long, the filaments 
filiform, glabrous, the anthers ovoid, obtuse, slightly brownish-punctate; sterile 
pistil subulate columnar, about 1.5 mm. long. Female flowers: calyx lobes, petals, 



each deeply bifid, filiform. Fruits subglobose, the valves broadly elliptic, about 


long and 2 mm. wide, the areolae distinct, yellowish-brown. 

Chiefly in thickets, at altitudes from 400 to 3,400 m.; China; flowering from 
April to May. 


hupei: Bo-moh-ping, Chun 3 570 (A); Changyang, Wilson 688 (A), 706 
(NY); Hsing-shan Hsien, Wilson 357* (A, type of Celastrus loeseneri Rehd. & Wils.; 
3H, MO), 560, 2309 (A, US), Chen logo (UC); Kui, Wilson 5. w., June 

Patung Hsien, Ho-chang Chow 

(A, US) ; without 

precise locality, Henry 315 (A, GH), 5909 (GH, NY, US), hunan: Hsikwang 
Shan, Handel-Mazzetti 583 (A); Hsinhwa, Handel-Mazzetti 8 JO (A); Wukang, 
Te-Hui Wang 1 14 (A), kwangsi: Lin-yuin Hsien, Steward & Cheo 429 (A, NY); Ling 
Wun, Lau 28207 y 28763 (A); Ling-yun Hsien, Lau 28558 (A), kwangtung: Suny, 
Wang 32148 (A) ; Yang-shan Hsien, Tsui 542 (A, MO, UC, US), kweichow: Anlung, 
Tsiang 9354 (A, NY); Kiang-kow, Tsiang 7499 (A, NY); Kweiting, Tsiang 5443 (A, 
NY), 5614 (A); Lang-tai, Yeh-tin, Tsiang 9513 (A, NY); Wai-ho, near Sio-chang, 
Tsiang 5601 (A); without precise locality, Cavalerie 496 (type of Celastrus cavaleriei, 
photo, A), sikang: near San-tao-chiao, 40 li from I-tung, Chiao 1720 (A); Konting 
(Tachien-lu), Wilson 4187 (A); Tienchuan, Tang 3424 (A) ,3486 (A, NY), szechuan: 
Chao-hua Hsien, Hopkinson 297 (S) ; Chengtu, Chien 5920 (A), Feng 19884 (A) ; Chin- 
tine Shan, Wilson 211 1 (A); inter Kuapie et Tahao-ko, Schneider 1 37 5 (A); Mian-ning 


[Vol. 42 

Hsien, Yu I 7 56 (A) ; Mow Hsien, Fang 5589 (A, NY) ; Nan-chuan, Ros thorn 1 57 2 (type, 
photo of Celastrus rosthornianus, A), 1753 (A, paratype of C. rosthornianus) ; Ning-yuen, 
Handel-Mazzetti 1305 (A); Telipu, Schneider 1131 (A); Tien-chuan, Yu-shih Liu 1319 
(A); Wen-chuan Hsien, Wilson 1 175 (A, US), 4159 (A); inter flumina Yalung et 
Nganning-ho, Handel-Mazzetti 2008 (US); Omei Shan, Chow 5707 (A), Chow 11742, 
12013 (A), Chu370l (A), Fang 16419, 17795 (A), Lee 2805 (A), 3069 (A, US), Wil- 

son 4781 (A); without precise locality, Faber 227 (NY), Henry $640 (A, GH, 

MO), 5734 (GH, NY), yunnan: A-tun-tze, Wang 69361 (A), Yu 7978 (A); Chung- 
tien, 2428, 2479, 3262 (A); La Kou, Maire 3633 ser. B (A, NY, UC); Lapping 
Hsien, Tsai 56096 (A); Liang-shan, Tsai 51340, 51343 (A); Likiang, Ching 22119 
(A) .Forrest 21201 (A, US), Rock 3885 (A, US), 8312 (NY, UC, US) ; Mar-li-po, Feng 
13124, 13369 (A); Mekong, Rock 6946 (A, UC, US); Mienning, Poshang, Yu 17893 
(A); Muli, near Lama-sery, Yu 14848 (A); Pan-pien-kai, Maire 7256 (UC) ; Pin-chuan 
Hsien, Tsai 52901 ( A) ; Ping-pien Hsien, Tsai 60781 (A) ; Si-chour Hsien, Feng 11720, 
12084 (A); Shun-ning, Wang 71910 (A); Suen-oui, Maire 81 (A); between Sung-kweh 
and Tengchuan, Schneider 2683 (A, GH, NY) ; Tso-si, Maire 254 (A) ; Wei-se Hsien, 
Tsai 57974, 5954$* 61753, 63059, 64147, 68293 (A); without precise locality, Forrest 
7812, 16226 (A) ; Tsai 57253, 57271, 57343, 57510, 57769, 60933 (A) ; Yu 5185, 7204, 
5$ 45 (A). 

This is a common species chiefly distributed in southwestern and central China. 
Since the leaf characters are quite variable, this species has sometimes been mis- 
identified. However, the characters of the flower, fruit, and seed are constant. 
The racemiform inflorescences bear clustered, usually sessile flowers; the disc-lobes 
are subquadrate, broader than long, and mucronate; and the seeds are ovoid or 
ellipsoid and usually yellowish-brown. These characters can distinguish Celastrus 
rosthornianus from related species. 

Leveille first identified a specimen collected by Cavalerie (no. 496) from 
Kweichow as Celastrus stylosa Wall., and later described it as a new species C. 
cavaleriei (non Leveille 1914) although he had previously published the same name 
based on another specimen. The earlier name subsequently was shown by Render 37 
to be a Myrsine. I have seen the type photo of the Cavalerie 496 specimen, and it 
compares closely with others collected in Kweichow. They are all fruiting speci- 
mens, and clearly belong to Celastrus rosthornianus. These Kweichow plants have 
grayish-brown, striate-fissured branches, and remarkably uniform, elliptic, firmly 
membranous leaves. This uniform population perhaps could be considered as an 

19. Celastrus punctatus Thunb. Fl. Jap. 97. 1784. (T.: Collector unknown, 
5. n. y A, photo!). 

Celastrus scandens sensu Thunb. in Trans. Linn. Soc. 2:332. 1794, non L. 
Celastrus punctulatus in Abh. Bayer. Akad. Muench. II, 4:150. 1845 (apparently an error 
for punctatus). 

Celastrus crispulus Regel, in Gartenfl. 9:407, tab. 312, f. 7-5. i860, ex char. &ill. 
Celastrus striatus Miq. in Ann. Mus. Bot. Lugd.-Bat. 2:210. 1865-66, non Thunb., ex char. 
Celastrus kiusianus Franch. & Sav. Enum. PI. Jap. 2:314. 1879, ex char. (T.: Savatier 
3528 ) . 

Celastrus orbiculatus var. punctatus (Thunb.) Rehd. in Bailey, Cycl. Am. Hort. 1:267. 

Celastrus articulatus Thunb. var. punctatus (Thunb.) Makino, in Bot. Mae. Tokyo 21:138. 
1907. * ' 

37 Jour. Arn. Arb. 15:292. 1934. 



Celastrus gracillimus Hay. Icon. Pi. Formosa. 5:24. 1915. (T.: Hayata s. n. A, photo!). 
Celastrus leiocarpus Hay. loc. cit. 5:22. 1915. (T.: Mori s.n., A, photo!). 
Celastrus longe-racemosus Hay. loc. cit. 5:23, pi. 3. 1915. (T.: Hayata s.n., TAI!). 
Celastrus geminiflorus Hay. loc. cit. 5:25, /. 9. 1915. (T.: Nagasama s.n., A, photo!). 
Celastrus elevativenus Hay. loc. cit. 6:14. 1916. (T.: Faurie s.n., A, photo!). 
Celastrus punctatus Thunb. var. microphyllus Li & Hou ex Hou, in Taiwania 1:172. 1950. 
(T.: Yamamoto 802, TAI!). 

Scandent shrubs, about 2-3 m. tall; branches terete, glabrous, hazel-brown, the 
lenticels rounded or oval, scattered and elevated; axillary buds deltoid, about 2 mm. 
long, the outermost bud scales sometimes sharply spinose. Leaves usually elliptic, 
the apex acute, the base usually cuneate, the margins remotely serrate, 2-7 cm. 
long, 0.8-3.0 cm. wide, membranous, glabrous, the primary lateral veins 4-5 pairs, 
slightly elevated below, immersed above, the veinlets obscure to distinct; stipules 
filiform, tufted, about 2 mm. long; petioles about 0.5-1.0 cm. long. Inflorescences 
axillary as well as terminal in the male plant, axillary only in the female, fascicu- 
late, racemiform, or solitary, the peduncles obsolete, glabrous, the primary peduncles 
suppressed to about 5 mm. long; flowers dioecious, white or pale green, the pedicels 
1.0-1.5 mm. long, the articulation at the middle or upper third of the stalk. Male 
flowers: calyx lobes open, deltoid to oblong, obtuse, glandular-ciliate, about 1 mm. 
long; petals usually oblong to slightly obovate-oblong, cilia te to slightly erose, 
2-5-4.5 mm. long and 1.3-1.5 mm. wide; disc cupuliform, the lobes erect, oblong, 
about half to one- third as long as the disc proper, obtuse; stamens arising from the 
margin of the disc proper, the filaments filiform, glabrous, the anthers broadly 
ovoid, obtuse, the base divided to above the middle, pink-punctate; sterile pistil 
subconical-columniform, slightly 3-lobed, about 1.5 mm. long. Female flower 
unseen. Fruits globose or subglobose, usually solitary, fasciculate, sessile or oc- 
casionally short-pedunculate, pale yellow when dried, the valves suborbicular, about 
6 mm. in diameter, 3- to 6-seeded; seeds broadly ellipsoid, about 2 mm. long, 
smooth, pinkish-brown. 

Hillsides and in thickets, at altitudes from 100 to 2350 m.; southeastern China, 
Riukiu Islands, and southern Japan; flowering from March to August. 

China: anhwei: Chu-hwa Shan, Ching 2661 (UC). chekiang: Changhua, Keng 
567 (A, UC) ; south of Siachu, Ching 1677 (MO, NY, UC, US), fukien: Amoy, Chung 

493 (SING), 1437, *447> 149* (A, UC), 151 5 (UC), 4604 (A, NY), 5957 (A); 

Diongloh, Ping En Chen 2447, 2701 (UC), Siu Ging Tang 13657 (UC) ; Foochow, Siu 
Ging Tang 1677, 7085 (A, UC) ; Kushan, Dunn 44 (A) ; Mt. Useke, Nagasawa s. n. (A, 
photo of type of C. geminiflorus). Taiwan: Arisan, Gressitt 1 41 (A, L, NY, S, U), 
185 (A, L, NY, S, U), Hayata s. n. y April 26, 1914 (TAI, photo of type of C. longe- 
racemosus; A), 5. n. (A, photo of type of C. gracillimus) ; Nanko-taisan, Sasaki s. «., July 
22, 1922 (A, photo); Kwarenko, Faurie s.n. (A, photo of type of C. elevativenus), 
Nakamura 3618 (TAI) ; Shinchow, Prov. Karenko, Wilson IIO96 (A, US) ; Taito, Mori 
5. n. (A, photo of type of C. leiocarpus) ; Mt. Taito, Yamamoto 802 (type of C. punctatus 
var. microphyllus, TAI). 

Japan: kyushu: Mt. Kirishima, Tashiro s.n., May 5, 1917 (A); Kudsi Kadsura, 
Tsuro Ome (?) 500 (A); Nagasaki, Oldham 161 (GH, L, S), 162 (GH), Wilson 6304 
(A); Satuma, Suzuki s.n., Aug. 13, 1930 (TAI); Tanegashima, Kita-tane, Mori s.n., 
Aug. 13, 1934 (TAI); Tanaga-shima, Wilson 6l2I (A). 

Riukiu Islands: Wright 54 (GH). 

[Vol. 42 


The type of Celastrus longe-racemosus is a flowering specimen whereas the 
type of C. punctatus var. microphyllus is a fruiting specimen. Until now no 
fruiting or flowering specimens, respectively, for these two species have been found. 
On further examination of additional specimens, it was observed that the fruiting 
branchlets of the type of C. punctatus var. microphyllus are the same as the so- 
called long-racemose inflorescences of C. longe-racemosus; consequently, they 
appear to belong to the present species because they cannot be distinguished from 
it. On comparing material from Fukien and southern Japan, this interpretation is 
confirmed by both morphological characters and geographical distribution. 

In general appearance, the present species differs conspicuously from Celastrus 
orbiculatus and more closely resembles C. rostbornianus. However, the deeply 
cleft, oblong disc lobes, the articulations located at the upper third or half of the 
stalk, as well as the distinct geographical distribution distinguish this species from 
related ones. 

20. Celastrus kusanoi Hayata, in Jour. Coll. Sci. Imp. Univ. Tokyo 30:60. 
1911; Icon. PI. Formos. 1:137. 1911; 5:20. t. 8. 1915, ex char. & ill. (T.: 
Kusano $. «.). 

Scandent shrubs up to 18 m. tall; branches terete, sometimes striate, brown, 
the branchlets glabrous to brownish-pubescent, both sparsely or rarely densely 
lenticelled, the lenticels orbicular or ovate; axillary buds deltoid, about 2.5 mm. 
long. Leaves broadly elliptic to nearly orbicular, rarely elliptic, the apex rounded 
to shortly cuspidate, the base broadly truncate, rarely obtuse to cordate, 5.0-10.5 
cm. long, 5-11 cm. wide, chartaceous, glabrous or pubescent on the veins below, 
the primary lateral veins 5-7 pairs, slightly elevated on both surfaces, the veinlets 
distinct below, obsolete above; petioles usually 1.5-2.5 cm. long, rarely up to 5 
cm. long; inflorescences axillary and cauline at the basal position of the current 
year's growth, rarely also terminal in the male plant, usually 3- to 7-flowered, the 
peduncles usually pubescent, sometimes yellow, the pedicels about 3-5 mm. long, 
usually pubescent, the articulation nearly toward the base of the stalk. Male 
flowers: calyx lobes deltoid, obtuse, entire, about 1 mm. long; petals obovate to 
oblong, rounded, about 4 mm. long and 1.5 mm. wide, ciliate, sometimes pubescent 
within on the lower portion; disc subfleshy, flat, the lobes obscure, truncate; 
stamens arising from the margin of the disc, about 3 mm. long, the filaments fili- 
form, papillose-tuberculate, the anthers ovoid, subcordate; sterile pistil conical, 
about 1 mm. long. Female flowers: calyx, petals, and disc as in the male; sterile 
stamens about 1.5 mm. long; pistil about 3.3 mm. long, the ovary subglobose, th 
style columnar, the stigmata 3-lobed, reflexed. Fruits globose or subglobose, the 
valves suborbicular or broadly elliptic, about 8-10 mm. long and 5-7 mm. wide, 
3- to 6 -seeded; seeds slightly lunate, minutely areola te, dark brown. 

Chiefly in thickets, at altitudes from 100 to 1,000 m.; southeastern China; 
flowering from February to April. 

CTO ^ H T T A ^ Bak v S u La Z 26l8 ± i 66<>6 (A l> Chin S-mai Hsien, Lei 3 OQ (A, NY, 
SING, UC, US); Fan Yah, Chun & Tso 44028 (A, F, NY, US), kwangtung: 




Fig. 10. Celastrus kusanoi Hayata 

Lokchong, Tsiang 1 393 (A, UC), Tso 20/88, 20796, 21026 (NY); Tsinleong Shan, 
Mei Hsien, Gressitt 1240 (A, MO) ; Lung-t'au Shan, Iu village, Kang-peng To 131 
(UC), 442 (MO, UC); Nam Shan, Ho-yuen Hsien, Tsang 28927 (A); Ying-tak, 
Kang-peng To & Kam-chow Wong 2784, 2928 (UC). Taiwan: Arisan, Faurie 1 374 (A) ; 
Kosyun, Kudo tf Suzuki 16048 (TAI), Suzuki 6195 (TAI) ; Kuaru, Yatnada 121 5 (TAI) ; 
Mt. Naier, Bansyoryo, Kawakatni & Mori 31 45 (TAI) ; Nanto, Wilson 998 1 (A) ; Taipei, 
Wilson 1 121 4 (A); Takou, Apes Hill, Henry 1893 (A, NY, TAI). 

Henry 38 first regarded the Formosan specimen, collected by himself (no. 18Q3) , 
as a variety of Celastrus articulatus Thunb., but did not give it a name. Later 
Hayata described a specimen collected by Kusano as Celastrus kusanoi Hayata, but 
he did not cite Henry's specimen. He stated that Celastrus kusanoi is near C. 
articulatus (C. orbiculatus) but differs from it in having more rounded leaves and 

38 A List of Plants from Formosa, p. 17. 1895. 



[Vol. 42 

carpels transversely wrinkled in the dried condition. His original description was 
based on a fruiting specimen; later 39 he obtained a flowering specimen, and then 
amplified his description. Celastrus kusanoi bears pubescent filaments, lunate seeds 
and usually nearly rounded leaves, hence it is easily distinguished. 


and sometimes even cordate leaves, whereas those from Hainan and Kwangtung 
usually bear suborbicular or broadly obovate leaves without cordate bases. 

Merrill and Chun 40 have said that Kwangtung records of Celastrus articulatus 
Thunb. (C. orbiculatus Thunb.) actually belong to C. kusanoi Hay. and that 
Thunberg's species does not extend as far as Kwangtung. Of the many Kwang- 
tung specimens labelled Celastrus articulatus which I have examined, none of them 
actually is that species; they are either Celastrus kusanoi Hay. or C. gemmatus Loes. 

21. Celastrus hirsutus Comber, in Notes Roy. Bot. Gard. Edinb. 18:233. 
1934, ex char. (T.: Forrest 17963). 

Scandent shrubs up to 12 m. tall; branches terete or slightly compressed, some- 
times more or less sulcate, densely brownish-pubescent or glabrescent, the lenticels 
elevated, large, ovate, elliptic or orbicular, rarely lacking on the current year's 
growth; axillary buds ovoid, usually about 3 mm., rarely 5 mm., long, the bud 
scales accrescent, sometimes persistent. Leaves broadly ovate to obovate, the apex 
round, shortly cuspidate, the base rotund or broadly cuneate, the margins crenate- 
serrate, 7-14 cm. long, 4-10 cm. wide, membranous, densely pubescent on both 
surfaces especially so on the veins when young; primary lateral veins 6-7 pairs 
prommently elevated below, slightly elevated above, the veinlets distinct below, 
immersed to obscure above; stipules laciniate, about 1 mm. long; petioles pubescent, 
about 1.5-3.0 cm. long. Inflorescences axillary and cauline at the basal portion of 
the flowering branch, usually 7- to 1 4-flowered, the peduncles hirsute, the primary 
peduncles 5-15 mm. long; flowers dioecious, creamy-white to greenish-yellow the 

pedicels 3 



Male flowers: 

» -* / »"« f«uc;>L.ciii. wnnouc, aDOUt 

1 mm. long; petals obovate-oblong, puberulous at the basal portions of both sur- 

mm. long and 2 mm 


the disc 

lobes, about 2.5-3.5 mm. long, the filaments subulate, papillose- tuberculate, the 
anthers ovoid, obtuse, cordate; sterile pistil small, ovoid, about 1.3 mm. long. 
Female flowers: calyx lobes, petals, and disc as in the male; sterile stamens about 
mm long; pistil 3-5 mm. long, the style columnar, the stigmata conspicuously 
3-lobed. Fruits subglobose, the valves broadly elliptic, about 8-10 mm. long and 


39 Icon. Pi. Formos. 5:20. 1915. 
40 In Sunyatsenia 5:111. 1940. 




Fig. 11. Celastrus hirsutus Comber 

In thickets, at altitudes from 1,400-2,500 m.; China, Burma, and Indo-China; 
flowering from March to April. 

Burma: Adung valley, 27°30' to 28°30' lat., and 97°30' to 98°30' long., Kittgdon 
Ward 9362 (A). 

China: yunnan: Kengma, Yu 17294 (A); Liukiang, Muchietu, Yu 2I0I2 (A); Ma- 
kwan Hsien, Tsai 51945 (A); Mar-li-po, Sze-tai-po, Feng 13765 (A); Mienning, Yu 
17952 (A); Ping-pien Hsien, Tsai 55421, 60261, 60499, 62396 (A); Shang-pa Hsien, 
Tsai 54799 (A); without precise locality, Forrest 17496 (A). 

Indo-China: Chapa, Tonkin, Petelot 5936 (A, NY). 

The particularly dense pubescence on both surfaces of the leaf makes Celastrus 
hirsutus easily recognized and distinguished from other species of the genus. On 

imilar to Celastrus kusanoi 

r has point 




[Vol. 42 


leaf and lenticel characters are distinct. For these reasons I am considering them as 

two separate species. 

22. Celastrus stylosus Wall 


Cat. 4313. 1831; Lawson, in Hook. Fl. Brit. Ind. 1:618. 1875 (pro parte); 



1904. (T.: Wallich 


branchlets, light brown to dark brown, sparsely lenticelled, the lenticels small, 

11 • -44 * 1 - - 


long. Leaves usually 
elliptic-oblong, ovate to obovate, the apex acute, the base acute or obtuse, the 
margins serrate, 6-1 5 cm. long, 3-9 cm. wide, membranous to firmly membranous, 
glabrous or rarely pubescent on the veins below, the primary lateral veins usually 
5-7 pairs, distinctly elevated below, plane or slightly elevated above, the veinlets 
prominent below, obscure to visible above; stipules filiform, about 1 mm. long; 
petioles usually 1-2 cm. long. Inflorescences axillary and cauline, usually 3- to 
7-flowered, at the basal part of the current year's growth (rarely also terminal in 
the male plant of Celastrus stylosus ssp. stylosus), distinctly pedunculate; th 
peduncles puberulous to glabrous, the primary peduncles 5-11 mm. long; flowers 
dioecious, green or pale green, the pedicels 2-5 mm. long, the articulation toward 
the base of the stalk. Male flowers: calyx lobes imbricate, oval to oblong, obtuse, 
slightly erose to entire, about 1.5 mm. long; petals obovate, obtuse, slightly erose, 

1.5 mm. wide; stamens arising between the disc-lobes, 
about 2.5 mm. long, the filaments filiform, fleshy, glabrous to papillose-tuberculate, 
the anthers ovoid, cordate; disc membranous, cup-shaped, the lobes distinctly 
arcuate or depressed- quadrate; sterile pistil about 1.5 mm. long. Female flowers: 
calyx lobes, petals, and disc as in the male; sterile stamens about 1 mm. lone: oistil 

2—4 mm 

bout 3 mm. long, the ovary 
J-lobed, each lobe bifid, flat 




;xed. Fruits subglobose, the valves 
10 mm. wide, 3- to 6-seeded; seeds 



A. Staminal filaments always densely papillose-tuberculate; the disc-lobes depressed- 
quadrate; leaves elliptic, membranous. India % i 
AA. Staminal filament, usually glabrous, rarely S^^^a^i^u^^' "^ 
lobes arcuate; leaves usually elhpt.c-oblong, firmly membranous. China..: 22b. ssp. glaber 

22 a. Celastrus stylosus ssp. stylosus. 

Celastrus neglecta Wall. Cat. 4341. 1831, nom. nud 

Gymnosporia neglecta Wall ex Lawson, in Hook. Fl. Brit. Ind. 1:619. 1875- Prain in 
Jour. Asiatic Soc. Bengal 73:198. 1904, Novic Ind 419 19ns (LLa r l \ 
neglecta WM. y Wallicb 434 i). 419 ' 19 ° 5 ' (based on Celastrus 

In thickets, at altitudes from 1,000 to 2,745 m.; India; flowering from March 
to July. ° 





Fig. 12. Celastrus stylosus ssp. stylosus 

1:62. 1936; in Contr. Bot. Surv. N. W 

India: assam: Khasi, Kurz 171 (CAL), Clarke 18678° (CAL). bengal: Darjeeling, 
Boswastore 73 (CAL), Clarke 26701* (CAL), Kurz s. n. (CAL). sikkim: Duphla Hills, 
Lister 1 57 (CAL) ; Lachen, Smith & Cave 964 (CAL) ; Moughoo, Lister s. n., March 1878 
(CAL) ; Punkabari, Lister s. n., April 1878 (CAL) ; Singolila forest, Rogers s. n., Jan. 1900 
(CAL); Sureil, Dr. Train's collector 439 (CAL). without precise locality: Anderson 
104 (CAL); Meebold 1 5912 (S). 

22b. Celastrus stylosus ssp. glaber Ding Hou, stat. et nom. nov. 

Celastrus hypoleucus forma 7. puberula Loes. in Engl. Bot. Jahrb. 29:445. 1900. (T.: 

Kosthorn /55<5\ A, photo!). 
Celastrus crassifolia Wang, in Chin. Jour. Bot. 

China 1:62. 1939. (T.: Yu 445> A!). 

Chiefly in thickets, at altitudes from 800 to 2,000 m.; China and Indo-China; 
flowering in April. 

China: anhwei: Kimen, Ip 41 (UC) ; Ching 3165 (UC) ; Wong Shan, Ling 1 1 04 
(UC). kwangst: Chuen Yuan, Tsoong 82041 (A); Kwei-lin Hsien, Tsang 38365 
(US) ; Ling-yuin Hsien, Steward & Cheo 87 (A, NY, S, SING), 396 (NY, S) ; Nanning, 
Seh-feng-dar Shan, Ching 8205 (NY, UC, US) ; Shang-sze Hsien, Tsang 24131 (A, MO) ; 
Pin-lam, Ko 55544 (A); Tsin-hung Shan, Hin Yen, Ching 7080 (A, US), kweichow: 
Mapo, Pingchow, Tsiang 6832 (A, NY), sikang: Tien-chuan Hsien, Tat & Feng 5238 
(A), szechuan: Nan-chuan, v. Kosthorn 1556* (A, photo of tvne of C. hypoleuca 


445 (A, type of C. crassifolia); Wa 

W — - 

(A) ; without 

precise locality, Henry 5559 (A, GH). yunnan: Chungtien, Feng 3365 (A); Hokin, 
Feng 747 (A); Lung-ling Hsien, Tsai 54563, 54578 (A); Mar-li-po, Chun*-dzia, Feng 
12727 (A) ; Mengtze, Henry 10^22, 11267 (A, MO) ; Ping-pien Hsien, Tsai 60420,60919, 
61156, 61629 (A); Shang-pa Hsien, Tsai 54929 (A); Shunning, Wumulun*, Yu 16592 
(A); Si-chour Hsien, Feng 12260, 12283 (A); without precise locality, Forrest 9396, 

15980 (K), Li 1 18 1 (A), Tsai 62841 (A). 

Indo-Chtna: Chapa, Tonkin, Vctelot 5829, 5946 (A, NY, US). 

These two subspecies are superficially similar except for the staminal filaments, 
the leaf shapes, and the geographical distributions as shown in the key. 

[Vol. 42 


Subspecies glaber is widely distributed in China; its extra-axillary inflorescences 
and firmly membranous elliptic-oblong leaves make it easy to separate from other 
related species. The staminal filaments are usually glabrous; however, a few speci- 
mens collected from Kwangsi and Yunnan are slightly papillose-tuberculate. 
Because of the intermediate leaf forms, staminal filaments, and distinctly geograph- 
ical distributions, I consider the Chinese population as a subspecies of C. stylosus. 

Under the provisions of the International Rules, priority of publication operates 
only within individual taxa. It would therefore be rather inappropriate to employ 
Loesener's epithet "puberula" in this instance, since Loesener chose the epithet with 
reference to the puberulence of the leaves of Kosthorn 1556*, which actually is an 
abnormal condition for the other known specimens of Celastrus stylosus ssp. glaber. 
Thus, I have chosen the epithet "glaber" with reference to the more significant 
character of the staminal filaments. 

20092, NY!). 

Jour. 13:37. 1934. (T.: Tsang 

cited, in part, non Prain. 



Celastrus oblanceifolia Wang & Tsoong, in Chin. Jour. Bot. 1:65. 1936, ex char. (T.: 
Tsoong 2443 ) . 

Scandent shrubs up to 10 m. tall; branches terete, glabrous, the young branch- 
lets sometimes brownish-pubescent, brownish-red to dark brown, both branches 
and branchlets lenticellate, the lenticels orbicular, sparse to dense. Axillary buds 
ovoid, about 2.5 mm. long, the outermost scales persistent, usually deltoid and 
spiny, acute to acuminate, accrescent, up to 5 mm. long. Leaves elliptic to ob- 
lanceolate, the apex acute, the base cuneate to obtuse, the margins remotely 
serrulate, 3-10 cm. long, 1.5-6.0 cm. wide, membranous, usually glabrous, rarely 
pubescent on the veins below, the primary lateral veins 4-5 pairs, arcuate toward 
the apex, immersed, distinct to slightly elevated below, obscure above, the veinlets 
usually obsolete on both surfaces; stipules laciniate, filiform, about 1 mm. long; 
petioles 7-12 mm. long. Inflorescences axillary and cauline at the lower part of 
the flowering branch, shortly pedunculate, the primary peduncles almost obsolete 
to 3 mm. long, usually 3 -flowered, puberulous; flowers dioecious, greenish-yellow, 
subsessile, the articulation just below the flower. Male flowers: calyx lobes im- 
bricate, ovate-deltoid to oblong, obtuse, entire, subglabrous without, accrescent, 
about 1.0-1.5 mm. long; petals oblong to oblanceolate, obtuse, slightly undulate, 
puberulent at the basal parts on both surfaces, about 3.5-4.5 mm. long and 1 mm. 
wide; disc fleshy, annular, entire; stamens arising from the margin of the disc 
proper, about 3 mm. long, the filaments filiform, densely papillose-tuberculate; 
sterile pistil ovoid, about 2 mm. long. Female flowers: calyx, petals, and disc as 
in the male; sterile stamens 1.5 mm. long, papillose; pistil subglobose, the style 
columnar, distinct, the stigmata discoid or slightly 3-lobed. Fruits subglobose, the 
valves broadly elliptic, about 8 mm. long and 7 mm. wide, 3- to 6-seeded; seeds 




Fig. 13. Celastrus aculeatus Merr. 

arcuate or semi-annular, about 4.5 mm. long and 1.5 mm. wide, brown, slightly 
wrinkled, distinctly areolate. 

In open fields or in thickets, from lowland up to about 900 m. elevation; south- 
eastern China; flowering from March to April. 

China: chekiang: Sia-chu, Ching 1664 (A, US); Taichow, Ching 1336 (US); Tsi- 
shun, Keng 285 (A); Tsingtien Hsien, Keng 73 (A), fukien: Amoy, Chung 1 514 (A, 

UC), 1693 (A, SING, US), 1773 (US), 4774 (A), 4777 (A), 5954, 6235 (A); Baek- 

liang and vicinity, Siu-Ging Tang 1 5609 (UC) ; Chuanchow, Chung 3085 (UC) ; Diongloh 
and vicinity, Ku Tai Lin 11677 (UC) ; Foochow, Norton 1 346 (UC), Chung-Chang 
Tang esf Shan En Ma 2936 (US); Kuliang, Chung 7264 (A, F); Kushan and vicinity, 
Cheng 1573 (UC); Kutien, Chung 4036 (A); Minhow Hsien, Chung 2065 (UC) ; Pu- 
cheng, Ching 2508 (A, UC, US) ; Tsze-chook-Hang, central Fukien, Dunn s. n. (HK, 
A) ; Yenping, Kuang-Han Chou 821 5, 8460 (UC) ; Yuen-fu gorges, Dunn s. n. (HK, A). 
hunan: Yi-chang Hsien, Tsang 23574 (A, US), kiangsi: Kaoan, Tsiang 10460 (NY); 
Kiennan Hsien, Lau 4367 (A, S, US); Lungnan Hsien, Lau 4801 (A, S, US); Lu Shan, 
Chiao 18768 (US); Swe-chuen, Hu 889 (A); Yung-shing, Hu 784 (A); without precise 
locality, Hu s.n. f 1920 (UC). kwangtung: Ho-yuen, Tsang 28843 ( A ); Kowloon, 
Wang 3060 (NY, SING) ; Mt. Lung-t'au, near lu, Kang-peng To et al 631 (US); Lung- 
tung, Tso 21640 (NY); Mei Hsien, Tsang 21416 (A, NY, S); Mui-nen Hsien, McClure 
& Shang 4 (UC) ; Naam-kwan-shan, Lung-men, Tsang 25268 (A) ; Sin-fung Hsien, Taam 


[Vol. 42 

724, 1041 (A); Kakchieh, Swatow, Gressitt 1767 (MO); Ta-ching, Chun 5539 (A); 
Tapu, Tsang 21189 (A, NY, S); Tsengshing Hsien, Tsang 20092 (NY, holotype of C. 
aculeatus; A, isotype) ; Wai-yeung, Tsui 154 (A, MO, NY, UC, US); Yao-shan, Sin 
U8S9 (NY). 

This species is well characterized by the arcuate seeds, the papillose-tuberculate 
filaments, the fleshy and annular disc, and the obsolete veinlets of the leaves. 
Celastrus aculeatus is a very distinctive species of southeastern China. Most of the 
specimens which are cited here have been identified as Celastrus hookeri Prain. 
Celastrus hookeri, however, is easy to separate from Celastrus aculeatus by its 
glabrous filaments, its cup-shaped disc, and its ovoid seeds. 

A specimen collected by Ching (2508) from northern Fukien has puberulent 
veins and slightly oblanceolate leaves. It matches the description of Celastrus 
oblancei folia Wang & Tsoong from southern Anhwei which also has oblanceolate 
leaves. The veinlets described for Celastrus oblanceifolia are reported to be obsolete 
here and the seeds are said to be curved, both typical characters of Celastrus 
aculeatus. Because of these morphological similarities and the identical geographical 
distribution, Celastrus oblanceifolia is placed in synonymy. 

24. Celastrus flagellars Rupr. in Bull. Acad. Sci. St. Petersb. II, 15:357. 
1857, ex char. 

Celastrus ciliidens Miq. in Ann. Mus. Bot. Lugd.-Bat. 2:8 5. 1865-66, ex char. 
Celastrus clemacanthus Levi., in Fedde, Rep. Spec. Nov. 8:284. 1910. (T.: Taquet 
632, A!). 

Scandent shrubs; branches and branchlets terete to slightly striate, glabrous, 
brown to dark brown, the lenticels small, oval, sparse, the young branchlets cling- 
ing by a line of filiform, branched aerial roots; axillary buds small, depressed-ovoid, 
usually protected by two prominent, broadly falcate, spiny scales, acuminate, about 
3 mm. long. Leaves broadly elliptic to suborbicular, the apex obtuse to shortly 



cm. long, 2-5 cm. wide, delicately membranous, glabrous or puberulous on the 
veins below, the primary lateral veins 4-6 pairs, distinctly elevated below, slightly 
elevated above, the veinlets distinct, sometimes elevated below, obscure above; 
stipules laciniate, filiform, about 5 mm. long; petioles 1.5-2.5 cm. long. In- 


the peduncles about 2-5 mm. long; flowers dioecious, white or yellowish- green, the 


upper half of the stalk. Male flowers: calyx lobes imbricate, oblong, obtuse, ciliate, 
about 2 mm. long; petals obovate-oblong to elliptic-oblong, ciliate to slightly 
erose, about 4 mm. long and 1.3 mm. wide; disc cup-shaped, the lobes incon- 


ments fleshy, filiform, glabrous, the anthers ovoid, obtuse, apiculate, cordate; sterile 

about 1 mm. long. Female flowers: calyx lobes, petals, and disc 



angular-cordate; pistil about 2.8 mm. long, the ovary subglobose, the style 



columnar, distinct, the stigmata trilobed, each lobe deeply bifid, reflexed. Fruits 
globose, the valves broadly elliptic to suborbicular, 6-7 mm. long and 5-7 mm. 
wide, 3- to 6-seeded; seeds shortly ellipsoid, about 3.5 mm. long and 2 mm. wide, 
brown, the areolae obscure. 

Lowland thickets, at altitudes up to 1,000 m.; China, Korea, and Japan; flower- 
ing from May to August. 

China: chekiang: Yu-tsien, Hu 1631 (A, UC). northeastern china: ad fl. 
Amur, Maack u *., 1855 (GH); sinus Possict, Maximowicz s.n., 1860 (GH). 

Korea: Chulla, Mrs. R. K. Smith s. n. (A) ; Gyouhfeng, Taquet 2723 (A) ; Quelpaert, 
scandens in muris agrorum Haouen, Taquet 632 (A, type of Celastrus clemacanthus) ; 
Hongo-san, Prov. Hogen, Wilson 10425 (A); Ping Yang, Jack s.n., Sept. 18, 1908 (A); 
Puk Han, Seoul, Jack s.n., Sept. 25, 1905 (A); Nam-san, Seoul (Heijyo), Prov. Heiki, 
Wilson 8450 (A, US). 

Japan: Kawagishi, Naganoken, Uno 21823 (A, NY); Musatre, Mitake, Hayakawa 
s.n., May 20, 1910 (S) ; Thinano, Suwa, Sakurai s.n., May 21, 1913 (A); Thinano, 
Nagano, Sakurai s.n., June 19, 1913 (A); near Lake Yamanaka, Dorsett & Morse 6lQ 
(A, US). 

This species is easily recognized by its characteristic ciliate-serrate leaves and 
the two persistent spiny outermost bud scales. Its branches sometimes produce 
aerial roots which function as attachment organs on smooth surfaces. The flowers 
are usually clustered on young shoots which occasionally elongate during the 
fruiting stage. 

Subgenus II. Racemocelastrus Ding Hou, subgen. nov. 

Frutices scandentes. Flores hermaphroditi; ovario triloculare, in quoque loculo 
ovulo singulo. Capsula semine singulo, ovulis manifeste abortivis binis. America 
centralis et australis. 

Type species: Celastrus racemosus (Reiss.) Loes. 


A. Inflorescences paniculiform, obviously compound. 

B. Inflorescences thrice- or multi-compound, up to 14 cm. long; lenticels dense. 

Mexico: Chiapas 25. C. lenticellatus 

BB. Inflorescences usually once or twice compound, up to 6 cm. long; lenticels 

C. Inflorescences usually clustered in the leaf axils; peduncles not associated with 
vegetative buds; leaves elliptic. Brazil, Venezuela, Costa Rica, British Hon- 
duras, Guatemala, El Salvador, and Veracruz, Mexico . 26. C. racemosus 

CC. Inflorescences solitary in the leaf axils; peduncles associated with vegetative 

buds; leaves ovate. Panama 27. G panamensis 

AA. Inflorescences racemiform, obscurely compound. 
D. Pedicels of the flowers 1-2 mm. long. 

E. Anthers distinctly apiculate; fruit valves about 12-14 mm. long, the septa 

1.0-2.5 mm. wide; leaves closely serrate. Central Mexico 28. C. pringlei 

EE. Anthers obscurely apiculate; fruit valves about 14-18 mm. long, the septa 
4.0-6.0 mm. wide; leaves remotely serrate to entire. Chiapas, Mexico; Guate- 
mala; Honduras; El Salvador 29. C. vulcanicolus 

DD. Pedicels of the flowers obsolete or less than 1 mm. long. 

F. Leaves oblanceolate-oblong; lenticels large, dense and elevated; anthers apiculate. 
Colombia 30 - C - caseariifolius 

FF. Leaves broadly elliptic or ovate; lenticels small, sparse and obscure; anthers 

not apiculate. Colombia and Venezuela 31. C. meridensis 


[Vol. 42 


Map 4. Distribution of seven species of Celastrm Subgenus racemocelastrus. 

25. Celastrus lenticellatus Lundell, in Bull. Torr. Bot. Club 67:616. 1940. 
(T.: Purpus 73?o, US!). 

Scandent shrubs; branches glabrous, black-brown, the lenticels small, dense, 
white, slightly elevated; axillary buds conoid, acute, about 2 mm. long. Leaves 
broadly elliptic, the apex abruptly short-acuminate, the base cuneate to rotund, 
the margins remotely serrulate, 10-20 cm. long, 5.0-7.8 cm. wide, chartaceous, 
glabrous, the primary lateral veins distinctly elevated below, plane or slightly im- 
pressed above, the veinlets prominent below, obscure to visible above. Stipules 

about 1 mm. long; pet 


axillary as well as terminal, solitary or fasciculate, thrice or multicompound, 
paniculiform, up to 14 cm. long, much branched at the base, the peduncles glabrous, 
the primary peduncles 1-5 mm. long; flowers bisexual, the pedicels about 1 mm. 
long, the articulation at the lower part of the stalk. Calyx lobes valvate, ovate, 

mm\ _*^ mm- ^ _ J --^ ^ _ — _ _ m— _ I _ _ ^\ P* *x ^*» 4 _-. _ 


oj r ~~* v ** & , *viuuu, V 

punctate, 1.8 mm. long and 1 mm. wide; disc fleshy, flat, about 




diameter, the lobes depressed-rectangular; stamens attached slightly under the 
margin of the disc, about 2 mm. long, the filaments filiform, glabrous, the anthers 
ovoid, obtuse, slightly apiculate; pistil short, ovoid, about 1 mm. long, the style 
stout, the stigmata obscure. Fruit unknown. 
Mexico: Chiapas; flowering in June. 

Mexico: Chiapas, Finca San Cristobal, Purpus 7370 (F, type; MICH; US, holotype). 

The paniculiform inflorescences and the densely lenticellate branches make this 
species easily separated from other Latin American Celastrus species. 

26. Celastrus racemosus (Reiss.) Loes. in Engl. Bot. Jahrb. 24:199. 1898, 

(as racemosa). 


astrus liebm 



Celastrus pachyrachis Lundell, in Lilloa 4:3 82. 1939. (T.: Jabn 476, US!). 
Celastrus mainsiana Lundell, in Lloydia 2:99. 1939. (T.: Lundell 6307, MICH!). 

Scandent shrubs up to 50 m. tall; branches terete, glabrous, blackish-brown, 
the lenticels scattered to dense, white, elliptic to orbicular; axillary buds conoid, 
acute, about 1 mm. long. Leaves elliptic to ovate, the apex acute, the base cuneate 
to rotund, the margins crenulate-serrulate, 5-12 cm. long, 3-5 cm. wide, firmly 
membranous, glabrous, the primary lateral veins 7-9 pairs, slightly elevated below, 
plane and visible above; stipules subulate, erose, about 1 mm. long; petioles 3-10 
mm. long. Inflorescences axillary, 1- to 3 -branched, once compound, up to 4 cm. 
long, the primary peduncle glabrous, about 3 mm. long; flowers bisexual, greenish- 
white, the pedicels 2-3 mm. long, accrescent, the articulation toward the base of 
the stalk. Calyx lobes imbricate, deltoid, obtuse, subentire, brownish-punctate, 
about 1 mm. long; petals oblong, obtuse, subentire, about 1.5 mm. long and 1 mm. 
wide; disc fleshy, flat, about 1.5 mm. in diameter, the lobes subreniform; stamens 
attached slightly beneath the disc margin, about 1.5 mm. long, the filaments linear, 
glabrous, the anthers ovoid, obtuse, rarely apiculate; pistil ovoid, about 1 mm. long, 
the style columnar and blunt. Fruits ellipsoid, the pedicels 2.5-5.0 mm. long, the 
valves broadly elliptic, 14-21 mm. long and 8-11 mm. wide, the septa 2.5-6.0 
mm. wide, 1 -seeded; seeds cylindric, about 10-13 mm. long and 7-9 mm. wide, 

pinkish-brown, areolae obscure. 

In wet forests or thickets, at altitudes 1,400-2,400 m.; Mexico, British Hon- 
duras, Guatemala, Costa Rica, Venezuela, and Brazil; flowering from March to 

Brazil: Rio de Janeiro, Riedel s.n. (US, type of Maytenus racemosus). 

British Honduras: El Cayo Dist., Lundell 6307 (MICH, holotype of C. mainsiana). 

Costa Rica: Heredia, Standley & Valerio 52024 (US). 

Guatemala: Tactic, Standley 71 349 (F). 

Mexico: vera cruz: near Jalapa, Pringle 8133 (K, MO, NY, S, UC, US); Mirador, 
Liebmann 1 48 7 1 (F, type of C. liebmannii; MO), 14872, 14873* H$74> *4875*> H^75 h 
(F), 14875 (GH, UC, US), Purpus 8926 (MO, NY, UC, US); Zacuapan, Purpus 7094 

(F, MO, NY, UC), 8080 (MO, NY, UC, US). 


[Vol. 42 

Venezuela: east of El Junquito, Steyermark 57015 (F, NY, US) ; Macarao, Jahn 476 
(US, holotype of C. pachyrachis) . 

This is a widely and disjunctively distributed species of the Latin American 
Celastrus. It is found in the rain forests or thickets from Vera Cruz, Mexico, 
southward to Rio de Janeiro, Brazil. 

Lundell 41 has stated, "from examination of the type photograph of C. racemosus 
(Reiss.) Loes., the similarity of that species and the type of C. liebmannii is 
striking," I have seen the types of Celastrus racemosus and C. liebmannii; in 
addition, I have examined specimens collected from Vera Cruz, British Honduras, 
Guatemala, Costa Rica, and Venezuela. They are all similar and excessively 
difficult to separate into two species. 

This species is closely related to Celastrus pringlei Rose and C. vulcanicolus 
Donn. It can be distinguished from them by the distinctly once compound, 
aggregate dichasia and longer pedicels. The other two species bear racemiform 
inflorescences and have very short pedicels. 

27. Celastrus 



Scandent shrubs; branches terete, glabrous, smooth, shining, reddish-brown, the 
lenticels small, obscure, elliptic; axillary buds conoid, acute, about 1 mm. long. 
Leaves ovate, the apex obtuse, the base rotund, the margins shallowly crenate, 
9-14 cm. long, 5.0-7.5 cm. wide, membranous, glabrous, the primary lateral veins 
7-8 pairs, elevated below, plane and distinct above, the veinlets slightly elevated 


11—13 mm. lone. Inflo 

usually twice compound, the primary peduncles glabrous, about 1.2-2.5 cm. long, 
associated with a vegetative bud in the axil; flowers bisexual, white, the pedicels 
about 1 mm. long, the articulation toward the base of the stalk. Calyx lobes im- 
bricate, obtuse, minutely erose, about 1.2 mm. long, brownish-punctate; petals 
oblong, rotund, more or less entire, about 2 mm. long and 1.2 mm. wide; disc 
fleshy, flat, about 2 mm. in diameter, the lobes subreniform; stamens attached 
slightly beneath the margin of the disc, about 2 mm. long, the filaments filiform, 
glabrous, the anthers ovoid, slightly apiculate; pistil short, conoid, about 1.5 mm. 
long, the style columnar and blunt. Fruit unknown. 

At altitudes 1,400-2,300 m.; Panama; flowering in April. 

Panama: chiriqui: Allen 31Q (MO, type; P). 

The floriferous branch of this species is very 


axillary vegetative bud. I have seen only the type collection. It is a young 
flowering branch. 

41 In Lilloa 4:3 80. 1939. 




Fig. 14. Celastrus Pringlei Rose 

28. Celastrus pringlei Rose, in Contr. U. S. Nat. Herb. 5:195. 1899. (T.: 
Pringle 6842, MOl). 

Celastrus longipes Lundell, in Lilloa 4:381. 1939. (T.: Palmer 106, MO!). 

Scandent shrubs up to 5 m. high; branches terete, glabrous, brown or reddish- 
brown, densely lenticelled, the lenticels elliptic or ovate, slightly elevated, white; 
axillary buds suborbicular, about 1 mm. long. Leaves narrowly elliptic, the apex 
acute to acuminate, the base attenuate, the margins serrulate, 6-10 cm. long, 2-4 
cm. wide, membranous, glabrous, the primary lateral veins 7 pairs, elevated below, 
plane and distinct above, the veinlets sliehtlv elevated beneath, visible above: 

bout 1 mm 

12 mm. long. Inflorescences 

long, the primary 

axillary, 1- to 4-branched, racemiform, usually 2.0-3.5 cm. long, the 
peduncles glabrous, obscure to about 5 mm. long, the secondary peduncles about 
3-8 mm. long, with 2-3 prophylls; flowers bisexual, white, the pedicels about 1.2 

long, the articulation usually at the upper half of the stalk. Calyx lobes 


imbricate, ovate, marginate, scarious and slightly ciliate, about 1.5 mm. long; petals 
obovate, rotund, slightly erose, 2.5 mm. long and 1.2 mm. wide; disc fleshy, flat, 
the lobes depressed-subquadrate; stamens attached just beneath the disc margin, 
about 2.5 mm. long, the filaments linear, glabrous, the anthers 


about 2 mm 

blunt. Fruits 


[Vol. 42 


septa 1.0—2.5 mm. wide, 1 -seeded; seeds broadly ellipsoid, 10—12 mm. long and 
7 mm. wide, pinkish-brown, shining, the areolae distinct. 

In forests, at altitudes 790-2,300 m.; Mexico; flowering from March to June. 

Mexico: canyons of mountains above Cuernavaca, Pringle 6842 (MO, type of C. 
pringlei; GH, L, S, UC) ; Temascaltepec, Hinton 290 (F), 3506 (GH, US), 3574 

(NY, US), 3717, 6076, 7203 (GH, US), 7380 (US), 9020 (GH, US), durango: 

San Ramon, Palmer 106 (MO, type of C. longipes; US), jalisco: McVaugb 10246, 
10308, 13906 (MICH), michoacan: Tancitaro, Leavenworth 60 Hoogstraal 1023 (F, 
GH, MICH, MO, NY), morelos: Sierra de Tepoxtlan, Pringle 6998 (GH, MICH, MO, 
NY, S, UC) . 

The type specimen of Celastrus longipes seems to have larger leaves and longer 
peduncles than the type of C. pringlei. These characters are variable in all the 
specimens I have examined, and I think that they might be due to the differences 
in altitude. I have found some floriferous branches which are subtended by foliage 
leaves. This subtending foliage leaf may be from the previous year's growth. For 
example, a specimen (McVaugh 10308), collected at altitudes 2,400-2,600 m. in 
Jalisco, Mexico, on April 14, has a main branch bearing three floriferous branches. 
Each of the branches is subtended by a foliage leaf. When I made a cross-section 
of the main branch two porous rings of spring wood showed clearly. From this I 
assumed it might be an evergreen species, but, Dr. McVaugh told me he has col- 
lected another specimen in fruit with all of the leaves fallen off. 

29. Celastrus vulcanicolus Donn. Smith, in Bot. Gaz. 61:373. 1916. (T.: 
Donn. Smith 2549, US! ) . 

Celastrus chiapensis Lundell, in Lilloa 4:380. 1939. (T.: Matuda 2080, MICH!). 
Celastrus siltepecanus Lundell, in Wrightia 1:155. 1946. (T.: Matuda 5192, MO!). 
Maytenus williamsii A. Molina R., in Ceiba 1:258. 1951. (T.: Merrill et al 15640, F!). 

Scandent shrubs up to 7 m. tall; branches terete, glabrous, gray to reddish- 
brown, the lenticels obscure on the gray branches while distinct on the reddish- 
brown ones, orbicular or ovate; axillary buds globose, about 1 mm. in diameter. 
Leaves elliptic, the apex acuminate or acute, the base cuneate or obtuse, the margins 
entire, repand, or slightly serrulate, 6-12 cm. long, 2.5-5.0 cm. wide, firmly mem- 
branous, glabrous, the primary lateral veins 7-9 pairs, curved toward the apex, 




flowered, the primary peduncles glabrous, obsolete to about 5 mm. long, the 
secondary peduncles about 3 mm. long, usually with 2 prophylls; flowers bisexual, 
pale green, the pedicels about 2 mm. long, the articulation usually at the middle 
or lower half of the stalk. Calyx lobes ovate or deltoid, thick, obtuse, marginate, 
scarious, slightly erose, about 1 mm. long; petals oblong-elliptic, obtuse, minutely 
erose, about 2.0-2.5 mm. long and 1 mm. wide; disc flesh v, flat, about 2 mm. in 


gin, about 2 mm. long, the filaments filiform, glabrous, the anthers subglobose, 
slightly apiculate; pistil 1.5 mm. long, the style slender and blunt. Fruits ellipsoid, 
the valves broadly elliptic, about 14-18 mm. long and 7-10 mm. wide, 1 -seeded; 



seeds ellipsoid, 12-14 mm. long and 6-9 mm. wide, reddish-brown, shining, the 
areolae obscure. 

In forests, at altitudes 1,300-2,400 m.; Mexico, Guatemala, Honduras, and El 
Salvador; flowering from December to January. 

El Salvador: Santa Ana, north of Metapan, Carlson goo (F). 

Honduras: Dept. Morazan, Merrill et al. 1 564.0 (F, type of May ten us williamsii; US) ; 
San Juancito, Williams d Molina R. 17100 (F). 

Guatemala: Dept. Alta Verapaz, Standley 71349 (F); Dept. Chiquimula, Steyer- 
mark 31480 (F); Dept. Quezaltenango, Steyermark 33635, 33692, 33775 (F), Standley 
65402 (F), 65411, 86972 (F, MICH); Dept. Sacatepequez, Standley 63667 (F), Donn. 
Smith 2549 (US, type of C. vulcanicolus) . 

Mexico: chiapas: Fraylesca, Siltepec, Matuda 5192 (MO, type of C. siltepecanus) ; 
Cascada, Siltepec, Matuda 5148 (F); Mt. Ovando, Matuda 2080 (MICH, holotype of 
C. chiapensis) , 3944 (F, MICH, MO, NY, US), 16396 (UC). hidalgo: Molango, Moore 
2698 (GH). oaxaca: Cumbre de Talea, Reko 4019 (US). 

This species is closely related to Celastrus pringlei Rose, but can be distinguished 
from it by the entire or remotely serrulate leaf margins, the obscure connectives, 
the larger fruits with wider septa, and especially the geographical distribution. 

The type specimens of Celastrus siltepecanus Lundell is similar to the present 
species except for the distinct and slightly elevated veins on both surfaces of the 
leaves, and the small distinct lenticels borne on the branches. These characters 
might be caused by the environment and are within the range of variation of the 

The specimens assigned to Maytenus williamsii A. Molina R. are Celastrus and 
are congruent with the present species. The leaves and fruits are smaller, which 
might be due to the high altitude of the habitat. 

30. Celastrus casearhfolius Lundell, in Lilloa 4:379. 1939. (T.: Lehmann 

Branchlets terete, glabrous, reddish-brown, densely lenticellate, the lenticels 


Leaves oblanceolate- 


oblong or elliptic-oblong, the apex acute to shortly acuminate, the base obtuse, 
the margins remotely serrulate, 4.5-10.5 cm. long, 1.4-4.0 cm. wide, chartaceous, 
glabrous, the primary lateral veins 7-9 pairs, veins and veinlets slightly elevated 
below, obscure above; petioles 3—5 mm. long. Inflorescences axillary, racemiform, 
1- or 2-branched, up to 2.5 cm. long, the primary peduncles obscure, the secondary 

long. Flowers (young) bisexual, the pedicels obsolete 
to about 0.8 mm. long, the articulation at the upper half of the stalk. Calyx lobes 
imbricate, ovate or ovate-deltoid, erose-ciliolate; petals ovate-oblong, obtuse, sub- 
entire; disc fleshy, flat, the lobes obscure, the stamens arising from the margin of 
the disc, the filaments glabrous, the anthers suborbicular, apiculate; pistil ovoid, 
the style short and columnar, the stigmata obscure. Fruits cylindric, the valves 
oblong, about 16 mm. long and 7 mm. wide, 1 -seeded; seeds cylindric, obtuse at 
both ends, 1 5 mm. lone and 6 mm. wide, black and smooth. 



[Vol. 43 


Colombia: Dept. de Antioquia, Daniel 329$ (F); highlands of Popayan, Lehmann 
s.n., 1,600-2,000 m. elev., May 1889 (F, holotype), 399 (L, NY). 

Celastrus caseariifolius is characterized by the innumerable, elevated and 
crowded lenticels, very much resembling a crowded colony of plant lice (aphids). 
Further, its chartaceous and oblanceolate-oblong leaves are distinctive. 

31. Celastrus meridensis Pittier, in Bol. Soc. Venez. Cienc. Nat. 3:423. 1927. 
(T.: Gehringer 298, US!). 

Maytenus meridensis (Pittier) Cuatr. in Fieldiana, Bot. 27 2 :82. 1951. 

Scandent shrubs; branches slightly striate, glabrous, brownish, the lenticels 
sparse, elliptic, slightly elevated, white, obscure on the young branchlets; axillary 
buds ovoid, about 2 mm. long. Leaves ovate, ovate-oblong, or obovate, the apex 
rotund to abruptly acute, the base cuneate to rotund, the margins remotely crenate- 
serrate, 4.5-7.0 cm. long, 2-5 cm. wide, firmly chartaceous, glabrous, the primary 
lateral veins 5-7 pairs, elevated below, plane and distinct above, the veinlets visible 
below, obscure above; stipules filiform, about 1 mm. long; petioles 3-7 mm. long. 
Inflorescences axillary, simple, racemiform, up to 5 cm. long, the primary peduncles 
obscure, glabrous, the secondary peduncles about 1 mm. long, with 2 small pro- 
phylls; flowers bisexual, white, the pedicels obscure, accrescent, up to 3 mm. long 
on the fruiting specimens. Calyx lobes imbricate, ovate, rotund, slightly erose, 
thick, about 1.2 mm. long; petals oblong, obtuse, slightly erose, about 2.6 mm. 
long and 1.3 mm. wide; disc fleshy, flat, about 1.6 mm. in diameter, the lobes 
depressed, subreniform; stamens attached just beneath the disc margin, about 2 
mm. long, the filaments linear, glabrous, the anthers ovoid, obtuse; pistil pear- 
shaped, about 1.5 mm. long, the style short, columnar and blunt. Fruits ovoid, the 
valves broadly ovate to suborbicular, about 11 mm. long and 7.5-9.0 mm. wide, 
the septa 2-3 mm. wide, 1 -seeded; seeds ellipsoid, 7 mm. long and 5 mm. wide, 
reddish-brown, shining, the areolae obscure. 

In thickets, at altitudes 2,490-2,700 m.; Colombia and Venezuela; flowering 
in July. 

Colombia: Cordillera Oriental, Dept. Boyaca, Cuatrecasas 1813, 1831 (F, US). 
Venezuela: Mucuruba, Gehriger 298 (US, type; F, NY). 

This species distinctly belongs to Celastrus. The morphological characters 
match the generic characters very well. The type specimen is a scandent, flower- 
ing plant. In addition to the type, I have several fruiting specimens at hand, 
which confirm this view. 

Cuatrecasas 42 transferred this species to Maytenus based on his own collections 


(1813 and 1831, F, US). He says, "My specimens were obtained from trees, 
justifying their inclusion in the genus Maytenus". On examining the specimens, 


scandent plant. 


Since all 

species are scandent shrubs, I assume that Dr. Cuatrecasas* specimens 
m trees" were in realitv scandent. 

42 In Fieldiana, Bot. 27 2 :82. 1951. 



Doubtful Species 

The type specimens or representative specimens of the following species are not 
available, while their original descriptions alone are not sufficient to place them. 
Celastrus discolor Levi, in Bull Geogr. Bot. 24:142. 1914 (T.: Cavalerie 

3919). China. 

Celastrus grenadensis Urb. Symb. Antill. 5:51. 1904 (T.: Eggers 6222). Ind. 

Celastrus microcarpus D. Don, Prod. Fl. Nep. 191. 1825 (T.: Kamroop s.n.). 
Reg. Himal. 

Celastrus racemosus var. trinitensis Urb. Symb. Antill. 5:52. 1904 (T.: 
Baptist e 58 57) . Trinidad, 

Celastrus repandus Bl. Bijdr. Fl. Ned. Ind. 1145. 1825 (T.: none). Java. 
Celastrus reticulatus Wang in Chin. Jour. Bot. 2:68. 1937 (T.: Leu 233). 

Excluded Species 


of which are species of Gymnosporia or Maytenus and require special study of 


Wall. Cat. no. 4342. 1831 = Chailletia gelonioides 

Hook. f. Fl. Brit. Ind. 1:570. 1875. 
astrus adenophylla Miq. Ann. Mus. Bi 
Thunb. Fl. Jap. 78. 1784. 

1865=Ilex crenata 

Celastrus alatus Thunb. Fl. Jap. 98. 1784 = Euonymus thunbergianus Bl. 

Bijdr. Fl. Ned. Ind. 1147. 1825. 
Celastrus alpestris Bl. Bijdr. Fl. Ned. Ind. 1145. 1825 = Perrottetia alpestris 
Loes. in Engl. & PrantL Nat. Pflanzenfam. Ill, 5:220. 1892. 

l Linnaea 15:458. 1841 = Acanthothamnus 
U. S. Nat. Herb. 23:684. 1923. 




1856, nom. nud. 

Wall. Cat. no. 4319. 1831 = Gymnosporia neglecta 

Hook. Fl. Brit. Ind. 1:619. 1875. 

Celastrus bilocularis F. Muell. in Trans. Phil. Inst. Vict. 3:31. 1859 = Maytenus 
bilocularis (F. Muell.) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 Abt. 

20b:135. 1942. 

Jack, in Malay. Misc. 1:19. 1820 


Cat. no. 4340. 1831. 

Celastrus boaria Baill. Hist. PL 6:26. 1877, in text, nom. nud. = Maytenus 

boaria Molina, Saggio Stor. Nat. Chile, ed. 1, 177. 1778. 
Celastrus bodinieri Levi, in Fedde, Rep. Sp. Nov. 13:263. 1914 = Ilex purpurea 
Hassk. Cat. PL Bogor. 230. 1844; Rehd. in Jour. Arnold Arb. 14:239. 1933. 
Celastrus buxifolia Wall, in Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind. ed. 


[Vol.. 42 

Carey & Wall. 2:396. 1824, in syn.: Celastrus rigidus Wall. = Gymnosporia 

wallichiana Laws, in Hook. Fl. Brit. Ind. 1:621. 1875. 
Celastrus cavaleriei Levi, in Fedde, Rep. Sp. Nov. 13:262. 1914 = Myrsine semi- 

serrata Wall, in Roxb. Fl. Ind. ed. Carey & Wall. 2:293. 1824; Rehd. in 

Jour. Arnold Arb. 15:292. 1934. 
Celastrus chungii Merr. in Sunyatsenia 3:253. 1937 = Tripterygium Wilfordh 

Hook, in Benth. & Hook. Gen. Pi. 1:368. 1862-67. 
Celastrus circumcissus Pavon, mss.; Briq. in Ann. Conserv. & Jard. Bot. Geneve 

20:253. 1919 = Maytenus orbicularis (Willd.) Loes. in Engl. Bot. Jahrb. 

50 (Beibl. 111):10. 1913. 
Celastrus colombianus Cuatrec. in Fieldiana Bot. 27 2 :81. 1951 = Ilex scandens 

Cua tree, in Lloydia 11:207. 1949. 
Celastrus confertus Ruiz & Pavon, Fl. Peruv. 3:7. 1802 — Maytenus confertus 

(Ruiz & Pavon) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 Abt. 20b: 146. 


Celastrus crenatus Forst. f. Prod. 19. 1786 = Gymnosporia crenata (Forst.) 
Seem. Fl. Vit. 1:40. 1865. 

Celastrus crenatus sensu F. Brown in Bull. Bishop Mus. Honolulu no. 130:158. 
1935, non Forst. = Gymnosporia sp., ex char. 

Celastrus crenatus sensu Hook. & Arn. Bot. Beechey Voy. 61. 1841 = Gymno- 
sporia vitiensis (A. Gray) Seem. Fl. Vit. 1:41. 1865. 

Celastrus crenatus Roth, Nov. PL Sp. 156. 1821 = Gymnosporia Montana 
Benth. Fl. Austral. 1:400. 1863. 

Celastrus crenulatus Wall. Cat. no. 4323. 1831, nom. nud. 

Celastrus cuneifolius (Wr. ex A. Gray) Gomez de la Maza in Ann. Inst. Segunda 
Ensefianza 2:172. 1895 = Euonymus cuneifolius C. Wright ex A. Gray in 
Mem. Am. Acad. n. s. 8:171. 1861 — not Celastrus (capsules 2-valved). 

Celastrus cunninghamii F. Muell. in Trans. Phil. Inst. Vict. 3:30. 1859 = May- 
tenus cunninghamii (F. Muell.) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 

2 Abt. 20b:136. 1942. 

flora F. M. Bailey in Queensland Agr. J 

29:178, pi. 22. 1912 = Maytenus cunninghamii (F. Muell.) Loes. in Engl. 
& Prantl, loc. cit. 1942. 

Celastrus dilatatus Thunb. in Trans. Linn. Soc. 2:332. 1794 = Orixa japonica 

Thunb. Nov. Gen. Pi. 3:57. 1783. 


1859 = Maytenus 

dispermus (F. Muell.) Loes. in Engl. & Prantl, Nat. Pflanzenfam. Abt. 2. 

20b:135. 1942. 

Celastrus diversi folia Hemsl. in Jour. Linn. Soc. 23:123. 1886 = Gymnosporia 

diversifolia Maxim, in Bull. Acad. Sci. St. Petersb. Ill, 27:459. 1881. 
Celastrus dubia Spreng. Syst. Veg. 1:774. 1825— not Celastrus (capsules 2-valved). 
Celastrus emarginatus Ruiz & Pavon, Fl. Peruv. 3:6, t. 229, f. a. 1802 = May- 



emarginatus Willd. Sp. PI. 1 2 :1128. 1798 = Gymnosporia emarginata 
Enum. PI. Zeyl. 409. 1864. 

Celastrus esquirolianus Levi. Fl. Kouy-Tcheou, 69. 1914 

Sieb. & Zucc. in Abh. Bayer. Akad. Muench. II, 4:146. (Fl. Jap 
1:38). 1845; Rehd. in Jour. Arnold Arb. 15:13. 1934. 




Celastrus esquirolii Levi, in Fedde, Rep. Sp. Nov. 13:262. 1914 = Sabia parvi- 

var. nitidissima Levi. Fl. Kouy-Tcheou, 379. 1915; Rehd. in 




Celastrus euonymoidea Levi. Fl. Kouy-Tcheou, 419. 1915 = Grewia feddei 
(Levi.) Burret in Notizbl. Bot. Gart. Berlin 9:678. 1926. 



pi. 280). 1896 = Gymnosporia sp., ex ill. 

13:263. 1914, quoad specim. 

Esquirol 3189 = Grewia henryi Burret in Notizbl. Bot. Gart. Berlin 9:674. 




1914, excl. Esquirol 31 89 

Grewia feddei (Levi.) Burret, loc. cit. 678. 1926. 
us finlaysonianus Wall. Cat. no. 4324. 1831, nom. nud. 

fournieri P; 

1841, nom. nud. 

f Panch & Sebert, Not. Bois Nouv. Caled. 234. 1874 = May- 
tenus fournieri (Panch. & Sebert) Loes. in Engl. & Prantl Nat. Pflanzenfam. 
2, Abt. 20b:138. 1942. 

Vahl, Symb. Bot. 2:42. 1791 = Elaeodendron glaucum 


Pers. Syn. PL 1:241. 1805. 
Celastrus haenkea Spreng. Syst. Veg. 4 2 :88. 1827 = Schoepfia flexuosa Roem. 

& Schult. Syst. Veg. 5:160. 1819. 
Celastrus hamelii Spreng. Syst. Veg. 1:774. 1825 = Rhamnus racemosus 

Duham. Traite Arb. et Arbust. ed. nov. 3:48. 1806. 
Celastrus heterophyllus Savi in Mem. Accad. Sci. Torino 38:163 (Mem. de C. G. 

Savi, p. 11, tab. II, /. 2.). 1835 — not Celastrus (fruit triangular). 
Celastrus heyneana Roth in Roem. & Schult. Syst. Veg. 5:421. 1819, p.p. = 

Gymnosporia heyneana Laws, in Hook. Fl. Brit. 1:620. 1875. 
Celastrus ilicifolius Schrad. in Goett. Gel. Anz. 1:716. 1821 = Maytenus 

truncatus Reiss. in Mart. Fl. Bras. II 1 : 5. 1861. 
Celastrus japonicus (Thunb.) Koch, Dendrol. 1:625. 1869 = Orlxa japonica 

Thunb. Nov. Gen. PL 3:57. 1783. 
Celastrus jodinii Steud. ex Gopp. in Gartenflora 3:312. 1854, in obs.: Ilex cunei- 

folia Hook. f. = Trichilia jodinii (Steud. ex Gopp.) Briq. in Candollea 

6:21. 1935. 
Celastrus kouytchensis Levi, in Fedde, Rep. Sp. Nov. 13:264. 1914 = Rhamnus 

crenatus Sieb. & Zucc. in Abh. Bayer. Akad. Muench. II, 4:146. (Fl. Jap. 

Fam. Nat. 1:38.) 1845. 
Celastrus leptopus Drake in Grandidier, Hist. Madag. 35. (Hist. Nat. PI. Atlas 

3, pi. 28o A ). 1896 = Gymnosporia sp., ex ill. 
Celastrus linearis var. madagascariensis Drake in Grandidier, loc. cit., pL 286 s . 

1896 = Gymnosporia sp., ex ill. 
Celastrus lineatus (Wr.) Gomez de la Maza in Ann. Soc. Espan. Hist. Nat. 19:239. 

1890 = Maytenus lineatus C. Wr. in Griseb. Cat. PL Cub. 54. 1866. 


[Vol. 42 


1904 = Gymnosporia sp. 

Wall, in Roxb. Fl. Ind. ed. Carey & Wall. 2:400. 1824— not 

Celastrus (ovary many-celled). 
astrus lycioides Bross. ex Willd. 
not Celastrus (branches spiny). 


Celastrus lyi Levi, in Fedde, Rep. Sp. Nov. 13:264. 1914 = Rhamnus esquirolii 

Levi, in Fedde, Rep. Sp. Nov. 10:473. 1912. 
Celastrus macrocarpus Ruiz & Pavon, FL Peruv. 3:8, /. 230, f.b. 1802 = May- 


Geneve 20:361. 1919. 


Celastrus ? magellanicus DC. Prod. 2:8. 1825 = Maytenus magellanicanus 

Hook. f. Fl. Antarct. 254. 1847. 
Celastrus mairei Levi, in Fedde, Rep. Spec. Nov. 13:264. 1934 = Sabia yunnan- 

ensis Franch. in Bull. Soc. Bot. Fr. 33:465. 1886; Loes. in Ber. Deut. Bot. 

Ges. 32:543. 1914. 


1798; Roem. 

& Schult. Syst. Veg. 5:428. 1819 — not Celastrus (leaves ternate). 
Celastrus maytenus Willd. Sp. PL 1 2 :1127. 1798 = Maytenus boaria Molina, 

Saggio Stor. Nat. Chile ed. 1. 177. 1778. 
Celastrus mexicanus Moc. & Sesse ex DC. Prod. 2:8. 1825 = Wimmeria mexi- 

cana (Moc. & Sesse) Lundell in Bull. Torr. Bot. Club 67:618. 1940. 
Celastrus micrantha Roxb. Hort. Beng. 86. 1814, nom. nud.; Fl. Ind. ed. Carey 

& Wall. 2:393. 1824 — not Celastrus (leaves pinnate). 
Celastrus mollis Decne. in Rev. Hort. II, 4:425. 1845-46 — not Celastrus (leaves 



PL Sp. 154. 1821== Gymnosporia Montana Benth 

1819; Roth, Nov. 

astrus may a O. Kuntze, Rev. Gen. 3 2 :37. 1898 = Moya spinosa Griseb. PL 

Lorentz. 63, pi. I, f. 3. 1874. 

astrus muelleri Benth. Fl. Austral. 1:399. 1863 = Maytenus sp. 

astrus myrtifolius Linn. Sp. PL 196. 1753 = Prunus myrtifolia (Linn.) 

Urban, Syn. Ant. 5:93. 1904. 

astrus neglect a Wall. Cat. no. 4341. 1831 = Gymnosporia neglecta Laws. 

Hook. FL Brit. Ind. 1:619. 1875. 

jm Wight 

1840 = Pittosporum flori- 

k 1834. 

Bot. Bemerk. 34. 1844 = Simmondsia 
Jour. Bot. 3:400, t. 16. 1844. 

Celastrus obtusi folia Roxb. Hort. Beng. 86. 1814, nom. nud.; Fl. Ind. ed. Carey 
& Wall. 2:393. 1824 = Gymnosporia trigyna Baker, Fl. Maurit. 50. 1877. 

Celastrus octogonus L'Her. Sert. 7. 1788 = Maytenus orbicularis (Willd.) 
Loes. in Engl. Bot. Jahrb. 50 (Beibl. Ill) :10. 1913. 



astrus op posit a Wall, in Roxb. Fl. Ind. ed. Carey & Wall. 2:398. 1824 
Pleurostylia cochinchinensis Pierre, FL For. Cochinch. Fasc. 20, sub. 
/. S°5y in text. 1894. 
astrus orbicularis Willd. mss. ex HBK. Nov. Gen. et Sn. PL 7:65. 


(Willd.) Loes. in Engl. Bot. J 


Celastrus orixa Sieb. & Zucc. in Abh. Bayer. Akad. Muench. II, 4:150. 1845 
Orixa japonica Thunb. Nov. Gen. PL 3:57. 1783. 

ilifolius Steud. Nom. Bot. ed. 2, 1:2 
Hook. Fl. Brit. Ind. 1:619. 1875. 

1840 = Gymnosporia ovata 

Celastrus ovatus Hill. Veg. Syst. 13:62, /. 12. 1824 = Colubrina ferruginosa 

Brongn. in Ann. Sci. Nat. I, 10:369. 1827. 
Celastrus ovatus Wall. Cat. no. 4308. 1831 = Maytenus ovata (Wall.) Loes. 

in Engl. & Prantl, Nat. Pflanzenfam. 2 Aufl. 20b: 140. 1942. 
Celastrus oxyphyllus Wall. Cat. no. 4312. 1831 = Gymnosporia acuminata 

Hook. Fl. Brit. Ind. 1:619. 1875. 
Celastrus pallidus Wall. Cat. no. 4307. 1831 = Gymnosporia Montana Benth. 

Fl. Austral. 1:400. 1863. 
Celastrus parviflorus Vahl, Symb. Bot. 1:21. 1790 = Gymnosporia sp. 
Celastrus parvifolius A. Rich. Ess. Fl. Cub. 349. 1845 — not Celastrus (fruits 2- 

valved and 1- to 2 -seeded). 
Celastrus pauciflora Wall, in Roxb. Fl. Ind. ed. Carey & Wall. 2:400. 1824— not 

Celastrus (ovary 1 -celled). 
Celastrus pentagyna Zipp. ex Span, in Linnaea 15:186. 1841, nom. nud. 
Celastrus quadrangulatus Schrad. in Goett. Gel. Anz. 1:716. 1821= Maytenus 

quadrangulatus (Schrad.) Loes. in Engl. & Prantl, Nat. Pflanzenfam. 2 

Aufl. 20b: 142. 1942. 
Celastrus retusa Poir. in Lam. Encycl. Meth. Suppl. 2:146. 1811 = Maytenus 

retusa Briq. in Ann. Conserv. & Jard. Bot. Geneve 20:351. 1919. 
Celastrus rhombifolius Hook. & Arn. in Hook. Bot. Misc. 3:170. 1833 = Iodina 

rhombifolia Hook. & Arn. ex Reissek, in Mart. Fl. Bras. 1 1 1 :78. 1861. 
Celastrus richardi Gomez de la Maza, Dice. Bot. Nom. Vulg. 25. 1889 = May- 
tenus buxifolius Griseb. Cat. PL Cub. 53. 1866. 
Celastrus richardi y. latifolius Gomez de la Maza in Ann. Soc. Espafi. Hist. Nat. 

19:239. 1890, nom. nud. 
Celastrus richardi 8. cochlearifolius Gomez de la Maza, loc. cit. 239. 1890 

Maytenus cochlearifolius Griseb. Cat. PL Cub. 53. 1866. 

Celastrus richardi c. elaeodendroides Gomez de la Maza, loc. cit. 239. 1890 
Maytenus elaeodendroides Griseb. loc. cit. 54. 1866. 

Celastrus rigida Wall, in Roxb. Fl. Ind. ed. Carey & Wall. 2:396. 1824 = Gymno- 
sporia wallichiana Laws, in Hook. Fl. Brit. Ind. 1:621. 1875. 

Celastrus robustus Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind. ed. Carey & 
Wall. 2:395. 1824 = Kurrimia sp. 

[Vol. 42 



1834 = Gymno- 

sporia rothiana Laws, in Hook. Fl. Brit. Ind. 1:620. 1875. 

astrus royleanus Wall. Cat. no. 4317. 1831 
loc. cit. 620. 1875. 

istrus rufa Wall, in Roxb. Fl. Ind. ed. Care 
sporia rufa Laws. loc. cit. 620. 1875. 


1824 = Gymno- 



14:242. 1933. 

1914 = Ilex macro- 
. in Jour. Arnold Arb. 

Celastrus seguini Levi, in Fedde, Rep. Spec. Nov. 13:262. 1914 = Myrsine semi- 
serrata Wall, in Roxb. Fl. Ind. ed. Carey & Wall. 2:293. 1824. 

Celastrus semiarillata Turcz in Bull. Soc. Nat. Mosc. 36 : :599. 1863 — not Celastrus 
(branches spiny). 

Celastrus sepiarius Dennst. Schlus. Hort. Ind. Malab. 31. 1818, nom. nud. 
Celastrus serrulatus Roth, Nov. PL Sp. 155. 1821, p. p. = Gymnosporia 

rothiana Laws, in Hook. Fl. Brit. Ind. 1:620. 1875. 
Celastrus spicatus Veil. Fl. Flum. 92, /. 138. 1827 = Gouania corylifolia 

Raddi, in Mem. Soc. Ital. (Modena) 18, /. 394. 1820. 

Celastrus spinifolius Larrafiaga, Escritos D. A. Larrafiaga 2:96. 1923— not 
Celastrus (capsules bivalved). 

Celastrus spinosus Royle, 111. Bot. Himal. 157. 1835 = Gymnosporia royleana 
Laws, in Hook. Fl. Brit. Ind. 1:620. 1875. 

Celastrus striatus Thunb. Fl. Jap. 98. 1784 = Euonymus altus (Thunb.) Sieb. 
Syn. PI. Oecon. Jap. 49. 1830. 

Celastrus suaveolens Levi, in Fedde, Rep. Sp. Nov. 13:263. 1914 = Ilex suave- 
olens (Levi.) Loes. in Ber. Deut. Bot. Ges. 32:541. 1914. 

Celastrus tetramerus Standley in Contr. U. S. Nat. Herb. 23:679. 1923 = 
Phyllanthus sp. 

Celastrus tristis Levi, in Fedde, Rep. Sp. Nov. 13:263. 1914 = Rhamnus napal- 


boaria Molina, Saggio Stor. Nat. Chile, ed. 1, 177. 1778. 


f. a. 1802 = Maytenus 


Jahrb. 29:446. 1900. 

1886 = Gymnosporia 

Celastrus verticillata Roxb. Hort. Beng. 18. 1814, nom. nud.; Fl. Ind. ed. Carey 

Wall. 2:391. 1824 
Fl. Ind. Orient. 154. 1834. 



/. b. 1802 = May- 





Celastrus vitiensis (A. Gray) [incorrectly ascribed to Benth. & Hook, by] Drake, 
Fl. Polyn. France, 30. 1893 = Catha vitiensis A. Gray, Bot. Phanerog., 

Wilkes U. S. Expl. Exped. 287, 
Gray) Seem. Fl. Vit. 40. 1865. 


Gymnosporia vitiensis (A. 

Celastrus wallichianus Spreng. Syst. Veg. 5, Index 150. 1828 = Gymnosporia 

wallichiana Laws, in Hook. FL Brit. Ind. 1:621. 1875. 
zstrus wallichianus Wall, in Roxb. Fl. Ind. ed. Carey & Wall. 2:4 
Gymnosporia wallichiana Laws, in Hook. loc. cit. 621. 1875. 


Celastrus tvallichiiG. Don, Gen. Syst. 2:8. 1832 = 

Linn. — not Celastrus. 
Celastrus wightianus Wall. Cat. no. 4332. 1831 


Pleurostylia cochin- 

chinensis Pierre, Fl. For. Cochinch. Fasc. 20, sub. /. 305 in text. 1894. 
Celastrus yunnanensis Levi. Cat. PI. Yun-Nan, 32. 1915 = Prema parvilimba 
P'ei in Mem. Sci. Soc. China 1:62. 1932; Rehd. in Jour. Arnold Arb. 15:324. 

Celastrus zeylanica Roth ex Roem. & Schult. Syst. Veg. 5:427. 1819 = Scutia 

commersonii Brongn. in Ann. Sci. Nat. I, 10:363. 1827. 

Enumeration of the Species 

Subgenus I. Celastrus 
Series 1. Paniculati 
1. paniculatus Willd. 


ssp. paniculatus 

ssp. serratus (Blanco) 

Ding Hou 




ssp. multiflorus (Roxb.) 

Ding Hou 
novoguineensis Merr. & Perry 
subspicatus Hook. f. 
richii A. Gray 
madagascariensis Loes. 
angulatus Maxim. 
scandens Linn. 

Series 2. Axillares 

8. monospermus Roxb. 

9. m&nospermoides Loes. 

10. hindsii Benth. 

11. glaucophyllus Rehd. & Wils 

12. hooker i Prain 

13. membranifolius Prain 

14. vanioti (Levi.) Rehd. 

15. hypoleucus (Oliver) Warburg 

16. gemmatus Loes. 

17. orbiculatus Thunb. 

18. rosthornianus Loes. 

19. punctatus Thunb. 

20. kusanoi Hayata 

21. hirsutus Comber 

22. stylosus Wall. 

22 a. ssp. stylosus 

22b. ssp. glaber Ding Hou 

23. aculeatus Merr. 

24. flagellars Rupr. 

Subgenus II. Racemocelastrus 

25. lenticellatus Lundell 

26. racemosus (Reiss.) Loes. 

27. panamensis Lundell 

28. pringlei Rose 

29. vulcanicolus Donn. Smith 

30. caseariifolius Lundell 

31. meridensis Pittier 


[Vol. 42 


Index to Exsiccatae 

Italicized numerals refer to collectors' numbers, s. n. (sine numero) to unnumbered 
collections; parenthetical numerals refer to the numerals of the species conserved in this 


Adams, J. W. 1 326, 1656, 286 5, 5016 (7) . 
Adams, J. W. & M. L. T. 1426, 21 31 (7). 
Adams, J. W., M. L. T. Adams & E. T. 

Wherry 1 519 (7). 
Adams, J. W. & G. Hopkins 907 (7). 
Adams, J. W. & R. Tash 568 (7). 
Adams, J. W. & E. T. WTierry 2 300 (7). 
Ahern's collector 2889 (Bureau of Sci. 

no.) (lb). 

Alcasid, — , & G. E. Edaiio 4.699 (lb). 
Alims, — . 1605, 2103 (9). 
Allard, H. A. 473, 5654 (7). 
D'Alleizette, Ch. 1288 (5). 
Allen, P. H. 319 (27). 

Anderson, T. 104 (22a); 107, 108, IOO 

Andrews, D. W. 39 (7). 

Backer, C. A. 37252 (lc). 

Bacon, E. S. s. n. (7). 

Baenitz, C. IIl8 (7). 

Balansa, M. 1870 (3). 

Ballard, C. A. s. «. (7). 

Barber, C. A. 6846, 7018 (la). 

Barnhart, J. H. 580, 1 194, 1394 (7). 

Bartlett, H. H. 6864 (9). 

Batchelder, C. F. s.n. (7). 

Bennett, W. M. 419, 1842, 2223, 5690, 

5720, 6308, 6902, 7923, 8682 (7). 
Bergman, H. F. 2101, 2630, 2855, 3120 

Bergman, S. 439 (17). 
Berkheimer, D. 4756 (7). 
Bicknell, E. P. 3836 (7). 
Biswas, K. 2033 (8). 

Bock, C. & A. v. Rosthorn 790 (6). 
Bohlin, B. 104 (17). 
Boswastore, — . 


— • 73 (22a). 
2032 (la). 
Brass, L. J. 12999 (9). 
Breitung, A. J. 6172 (7). 
Brooks, W. P. 44, 544 (17). 
Bush, B. F. 118, 732 (7). 
Bush, Y. 60 (7). 

Calder, C. C. & M. S. Ramaswami 6*d 

(la). ^ 

Carr, C. E. 11260 (3) ; 13036 (9) ; 15698 

Carlson, M. C. goo (29). 
Carter, H. G. 466 (lc). 

Cavalerie, J. 406 (IB) ; 3976 (10). 
Chandler, H. P. 510 (7). 

Chase, V. H. 255, 853, 4057, 16652 (7). 
Chen, P. E. 2447, 2701 (19). 
Chen, Y. IOOO (18). 

Cheng, C. H. J 57 3 (23). 

Cheo, T. Y. & L. Yen 1 03 (17). 

Chiao, C. Y. 1720 (18); 1721 (11); 21 16, 
2126, 2395, 2612, 3016 (17); 1 8641 
(22b); 18768 (23). 

Chiao, C. Y. & C. S. Fan 458 (10); 767 

Chien, S. S. 5920 (18). 

Ching, R. C. 1 336, 1664 (23); 1677 
(19); 2508 (23); 2638 (16); 2661 
(19); 2700 (11); 2845 (22b); jodj 
(15); 316S, 546l (22b); 6057 (14); 
7080 (22b); 7/20, 7525 (8); 7566 
(16); A?05 (22b); 8433 (8); 20^p, 

20<5jc? (11); 20682 (16); 2/67o > , 220^2 

(11); 22II0 (18). 
Ching, — . & — . Tso 580 (17). 
Chou, K. H. 8215, 8460 (23). 
Chow, C. L. 5707 (18); 6448, 6468 (11). 
Chow, H. C. 315 (18); jdp (17); 405, 

495 (6)1548 (10); 572 (18); 597 (6); 

7<?5 (10); 901 (16); J^J(5 (11); Jp^ 

(10); 2005 (6); 7597 (11); 8639 (10); 

8028 (11); ppo<5 (16); 1/7/2 (10); 

1 1 742, 12013 (18). 
Chu, K. L. 3701 (18); 3707 (16). 
Chun, N. K. 40143, 40276 (8); 42959 


Chun, N. K. & C. L. Tso 44028 (20); 

44165, 44169 (8). 
Chun, W. Y. 3/07 ( 8 ) ; 3503 ( 17) ; 3570 

(18); 3724 (6); 3<?oV (16); ^odd (6); 

4427 (15); 422p (16); 4831 (8); ^0 

(10); ,#43 (8); 5013 (16); 553P (23). 

Chun, W. Y. & S. S. Chien 8050 (6). 
Chung, H. H. 335 (10); 493 (19); 954, 

IO84, 1 174, I206 (10); 1437, J 447, 
H9I (19); 1514 (23); 15/5 (19); 1693 
(23); 777/ (10); 7773, 2065, 308 5, 
4036 (2i);4599 (10); 4604 (19);4774> 

4777 (23); 47<?3, 4851 (io); 534/ 

(19) ; 5601 (10) ; 5P54 (23 ) ; 5957, 5977 

i 19 l l o 62 } 2 (10 > ; 62 35> 7264 (23); 
8436, 8446 (10). 

Chung, H. H. & S. C. Sun 592 (16). 

Chung, Z. S. 83605 (16). 

Clarke, C. B. 682c (12); 56/5 (10); 

/20*<5 (22a); 18678C (22a); 23<5o7B, 

2^57, 26339 (la); 2670/B (22a); 

27O40D (12); 27683 (8); 3575^ (12); 

37949B (8); £rp//C (10); 45/52 (8); 

46569 (12). 
Clemens, J. & M. S. Clemens 3523 (2); 

458l, 5084 (9); 5/52, 5394 (2); 6253 






31827, 32052, 32120, 32860, 33182, 

50253, 50641, 51295 (9). 
Clemens, M. S. 11073 (2); 11680 (7); 

43683 (3). 
Clokey, I. W. 2479 (7). 

Coert, J. H. 3758 (lb). 

Colani, E. 4049 (lc). 

Collect, H. 588, 672 (la). 

Corbett, — . & — . Williams 88 (7). 

Core, E. L. 3778 (7). 

Cory, V. L. 14152 (7). 

Cuatrecasas, J. 1813 (31). 

Cuiran, H. M. 7499 (lb). 

Cuming, H. 1209, 1 324 (lb). 

Curran, H. 3415 (lb). 

Curtis, C. 404, 2005 (9) . 

Daniel, H. 3295 (31). 

Darsteri, A. C. 1 7 (8). 

David, A. J 78 3 (17). 

Davis, John. 199, 716, 720, 3709, 6368, 

7344 (7). 
Day, M. A. 43, 96 (7). 

Deam, Chas. C. 15456, 18042, 18240, 

18331, 18810, 19320, 35238> 379H, 
38 1 41 (7). 

Degener, Otto 1 4798, 14940, 14973, 14978, 

15050, I53I5> J5330, 1*3*2, 15424, 
*5427 (4). 
Demaree, D. 6746, 6955, 11340, 11572 

Deplanch, — . 92 (3). 
Desiloa, F. s. n. (8). 
Dickason, F. G. 5845 (la). 
Dickey, S. S. 27 (7). 
Dorgelo, J. D. 2290 (lc). 
Dorsett, P. H. & W. J. Morse 619 (24); 

1070, 7074, 7152, 7206 (17). 
Dougboo, — . 86225 02). 
Drummond, J. R. 21515, 21516, 21651, 

21691 (la). 

Ducloux, F. 132 (16). 
Dudgeon, W. & L. A. Kenoyer 237 (la). 
Dunn, — . 44 (19); 2479, 2479B, 3420 

Duthie, L F. 821 1 (la). 

Eccles, R. G. 29 (7). 
Edafio, G. E. 999 (9); 77599 (lb). 
Edmondson, T. W. 3426, 3876 (7). 
Eggleston, W. W. & K. M. Wiegand 22503 

Elizabeth, E. E. 351 1, 3903 (7). 
Elmer, A. D. E. 11411 (9). 
Elwell, L. H. 202, 203 (7). 
Emig, W. H. 370 (7). 
Endert, F. H. 3846 (9). 
Enequist, J. 435 (7). 

Engelmann, G. 1107 (7). 

Esquirol, J. 4 O^). 

Evans, I. H. N. 1 3247 (lc). 

Evers, R. A. 143 (7). 

Evers, R. A., G. N. Jones & F. F. Jones 

541 (7). 
Eyles, D. E. & M. S. Eyles 

Eyles, M. S. 8380 (7). 

8505 (7). 

190 (22b); 210, 

Faber, E. 227 (18). 
Fan, C. S. & Y. Y. Li 

236 (11); 263 (22b). 

Fang, W. P. 571 (6); 595 0«); ^02* 

(11); JJ<52 (6); 2124 (16); 3089 (11); 
3424, 3486 (18); J7J7, ^//d (11); 5529 

(16); 55<?9 (18); 12883, 13051 (H)s 

/4pc?J (10); 752// (22b); 75227 (10); 
752<fo (22b); 76479 (18); j<56?4, 77377 

(11); 17795, 19884 (18). 
Fang, W. P. & T. C. Lee 3575 (16). 

Faurie, U. 1374 (20); 1376 (10); 2502, 
d/JJ, 6134 (17); <?OJ5 (19). 

Fay, M. J. 208 (7). 

Fender, F. S. 54J, 55<5, d/5, 616, 863, 
1 1 58, 1329, 141 1, 3264 (7). 

Fendler, A. 104 (7). 

Feng, K. M. 747 (22b); 787, 1226 (11) 
2^2<?, 2479 (IS); 2868 (16); 3047 (11) 
J2<52 (18); 3299 (11); Jjtfj (22b) 

1 1 445, 11530 (8); //d<?5 (10); 1 1 720 
(18); 77#JO (8); 77<5#2 (10); 12084 
(18); 12260, 12283 (22b); 12469 (8); 
72727 (22b); 13068 (10); 7J72^ (18); 
7J27S (10); 13369 (18); 7J5<?7, 7J454, 
13464 (10); 73472, 7J749 (8); 1 3765 

(21); 7J<?47 (10). 
Fenzel, G. 502 (17); 67^, 7J<? (6); 927 

(15); 2<?77 (6); 2827 (17). 
Fernald, M. L. 167,2709 (7). 
Fernald, M. L. & B. Long 9858, 9859, 

12716, 13680, 13973, 1 4031 (7). 

Fester, G. S. 1 288 (7). 
Fischer, C. E. C. 4739 (la). 
Fogg, J. M., Jr. 5807, 8605, 7410, 8213, 
8554, 10269, 12293, I2 5 2 7> 13032, 14 6 78, 

14835 (7). 
Forbes, H. O. £27, 961, 980, 980a (10); 

3829 (22a). 

Forrest, G. 7812 (18); 8704 (16); 9396 

( 22b) ; 9948, IO069, IOI42, III52 (11); 

7777J (16); 77<?74 (8); I205O, I45W 

(16); 15980 (22b); 16226 (18); 16241 
(16); 16538 (11); 17496 (21); 27207 

(18); 27^ (16); 22235 (")i »J" 
(16); 22475, 2J2J0 (11); 26407 (8). 

Fosberg, F. R. 15146 (7). 
Fox, R. B. 4864 (lb). 



[Vol. 42 

Friesner, R. C. 

Fung, H. 

II7H (7). 
OOl6l (8); A-551 (10). 

Galen, J. 480 (7). 

Gamble, J. S. 8780 (la). 

Gamble, N. IQ26A (12). 

Gammie, G. A. 1 1 79 (12). 

Gates, — . 7468 (lb). 

Gates, F. C. 15880, 18744 (J). 

Gehriger, — . 208 (31). 

Gilbert, F. A. 187 (7). 

Gilbert, F. A. & C. R. Gilbert 285 (7). 

Gilbert, K. S. 6 (17). 

Gillespie, J. W. 4182 (A). 

Gleason, H. A. 9213 (7). 

Gleason, H. A., Jr. 639 (7). 

Goering, — . 88 (19). 

Grant, M. L. 2p7J (7). 

GrassI, C. O. 2250, 3218 (7). 

Greenman, J. M. 783, 837, 2409, 2510 

Greenman, J. M., O. E. Lansing, Jr. & R. 

A. Dixon 44 (7). 
Greenwood, W. 900 (4). 
Gress, J. L. 2018 (7). 
Gressitt, J. L. 141, 185 (19); 1240 (20); 

1767 (23). 

Griffith, — . 605 ( 1 2 ) ; d&5, /002 ( g ) . 

Groff, G. W. 6132 (lc). 
Grood, J. H. 799 (7). 

Hallier, — . 462 (10). 

Hance, H. F. 370, 22191 (10). 

Handel-Mazzetti, H. 132 (6); jtfj, <?/o, 

/505, 2000* (18); 5d$>2 (6); /307 (16). 

Haneeflf, Md. & Md. Jan 2446 (9). 

Harbison, T. G. 1611 (7). 

Harsukh, — . 21 400 (la). 

Hayden, A. 9590, 10070 (7). 

Hay ward, H. E. 616 (7). 

Heiser, Charles B., Jr. 1801 (7). 

Heller, A. A. & E. G. Halbach 522 (7). 

Henrici, S. II 5 1 (7). 

Henry, A. 0986a (11); 1893 (20); 2084 

(6)12837 (15); 31 1 5 (IS); 3241 (10); 

54<>5 ( 6 ) ; J4P5, J495A, j<?5<5 ( 1 ) ; 3883 

(6) ; 5559 (22b) ; 5640, 5734 (m ; 5^7 

(15); 5009 (18); 5925* (6); 5Q 86 
(11); 5p<?<5* (18); 677/, 68ll (15); 

7<5/4 (16); pd/p, 9679 A (11); p 7( ?2, 

P7<?2A (16); I0446 (8); J0522 (22b)- 

J05J/ (16); 7055p (10); 10955 (8);' 

II267 (22b); I/39P ( 8 ); //^j (16) 
II572 (la); J/p/2 ( 8 ); 11972A (lc): 
//P72B (8); II972C (10); 11993] 

1 1 993 A, 1 21 22, 1 21 22 A, 12122B (lc)- 

12572 (la); 1 2 572 A, 1 2572 B ( 1 c ) . 

Hermann, F. J. /<?20 (7). 

Hers, J. 26 (17); 2dp (6); J45 (17); 
4/7 (6)5 619, 704 (17); 772 (6); 828 
(17); <&>/ (6); 7272 (17); 1288, 1314, 
1712 (6); 1758 (17); J«?J/, 2^2, 2670 

(6) ; 27p/, 27P5 ( 17) ; 2pj2, 3010, 3097 

Hieronymus, P. 443 (la). 

Hinckley, L. C. UOc? (7). 

Hinton, G. B. 2po, 3506,' 3574, 3717, 

6076, 7230, 7380, 9020 (28). 
Hitchcock, A. S. 60 (7). 
Ho, H T. 60086 (8). 
Hooper, D. 38801, 39537 (la). 
Hopkinson, — . 297 (18). 
Home, J. 1 135 (4). 

Horr, W. H. & R. L. McGregor E493 

Horsey, R. E. 2po, 1283, 1488, 1545, 1581, 

1748, 1754, 2258, 2380 (7). 
How, F. C. 70368 (10); 70573 (lc); 

73564 (8). 
How, F. C. & N. K. Chun 70273 ( 8 ) . 
Hsiung, Y. K. 5516 (23); 5655, 570 1 

(16); 57^ (23). 
Hu, H. H. 784, 889 (23); 1 1 19, 1585 

(16); 1631 (24). 
Hu, W. K. 7380 (11). 
Huk, A. (52 (la). 
Hunnewell, F. W. 5<5<?5 (7). 
Hurusawa, I. 10650 (17). 

Ip, N. K. 4/ (22b). 

Jack, J. G. 3857{7)- 

Janet, — . A7p, A348, A349, 1639 (17). 

Jenkins, P. H. & G. L. Stebbins, Jr. 170 

John, A. 476 (27). 
John, H. S. 735, 2795 (7). 
Johnson, F. W. 249, 1867 (7). 
Jones, G. N. 11456, 13861, 16251 (7). 
Junghuhn, — . 27, 1 05 (10). 

Kanehira, R. & S. Hatusima 72/02 (9). 
Kang, — . i 7 4 (17). 

Kanjilal, Upendranath 1801 (8): 4076 

(10). * y ' 

Kasin, — . 547 (ic). 

Kawakami, T. & U. Mori 3145 (20). 
Keith, — . 2 13 (213). 
Kellogg, J. H. jo'p (7). 

Keng Y L. 75, 26*5 (23); 576 (19). 

Kerr, A. F. G. 9941 (10). 

Khalil, A. j. n. (la). 

Khan, B. // (lc); 36,259 (la). 


J702J (7). 




Kimura, Y. S°7 7 )- 

King, G. s.n. (la); 2134 (8). 

Dr. King's collector s.n. (la); 178, 233 

(&);454, 479 (la); 5690, 6928 (9). 

Kirk, M. D. 307 (7). 
Kjellberg, G. 2311 (9); 2907 (lb). 
Ko, S. P. 50089, 31463(8); 55544 (22b); 
56089 (10). 

Koelz, W. 4266, 4377 (la). 
Koorders, S. H. 19632P (lb); 27372P, 
29803P (lc); 301 78 P (lb); 424 J<?/3 

Korthals, — . 747 (10). 
Kostermans, A. 441, 1027 (lc). 
Krotkov, P. V. 7587, 9197 (7). 
Kudo, Y. & S. Suzuki 16048 (20). 
Kuhl, — . & Van Hasselt 18 (lc). 
Kurz, S. 65 (la); 171 (22a); 182 (8); 

7926" (la); 1927 (lc). 

Lace, J. H. 2229 (8); 5/50, J2<?0 (la); 

5753, 5775 (12). 

Lakela, O. 3018 (7). 

Lam, H. J. 2466 (lc). 

Lambert, B. B. 12 (7). 

Lansing, O. E., Jr. 3525 (7). 

Lau.S.K. 283 (lc); 693 (\0);8oo (8); 

2(572 (10); 2750 (8); J/pJ, 3658 (lc); 

57/7 (8); 4367,4801 (23); 26184 (20); 

(10); 2<52<?5 (8); 26606 (20); 

2<5<52^ ( 8 ) ; 2<?207, 28558, 28707, 28763 

(18); 20043. 
Laybourn, W. A. 4 (7). 
Leavenworth, W. C. & H. Hoogstraal 1023 

Lee, T. C. 2805, 3069 (18) ; J/<5<? (22b) ; 

3561 (11); 4557 (10). 
Leeuwen, W. M. Docters v. IIO98 (2). 
Lehmann, — . 399, s.n. (30). 
Lei, C. I. 129 (10); 309 (20); 440 (10). 

Levine, C. O. 1 80, 240, 441, 484, 1 2 36 

(CCC nos., 10). 
Li, C. F 77769 (17). 
Li, M. K. Il8l (22b); 2807, 2947 (lc). 
Li, T. F. 69 (17). 
Liang, H. Y. 62583 (8); 63042 (lc); 

63419 (8); 63892 (lc); 65293 (8). 
Licent, E. 7J202 (17). 
Liebmann, — . 14871, 14872, 14873, 

14874, 14875, 14875a, 14875b (26). 

Lin, Y. T. 1 1677 (23). 
Ling, K. 7759 (16); 7794 (22b); 72^<5, 
1349 ( 16) ; 7402, 7607, 1664, 2197 ( 17) . 
Liou, K. M. 385 (17). 
Lippincott, C. D. 26 (7). 
Lisi, A. G. 30 (7). 
Lister, J. L. 68 (10); 757 (22a). 

Liu, J. C. 398, 1 176 (17). 

Liu, Y. & C. W. Wang 82947 («)• 

Liu, Y. S. 7747 (16); 1184 (22b); 7J79 
(18); 1423 (16); 7774 (22b); 

Lloyd, C. G. 480 (7). 

Long, B. 3498, 5700, 8273, 9798, 13988, 
14264, 14380, 15412, 15616, 16886, 
17750, 18211, 18231, 18743, 18750, 
19331, 19554, ^9747, 20887, 20921, 
21818, 22802, 23299, 24117, 25248, 
26601, 26661, 27032, 27965, 28195, 
29256, 20891, 30948, 31319, 31656, 
31910, 33202, 33250, 33572, 34655, 
35526, 37261, 38396, 38632, 38876, 
39362, 40005, 40225, 42511, 48957, 

2, 53034, 53957, 54i62, 54300, 

54659, 55338, 55447, 55517, 56020, 
59108, 62352, 62569, 64027, 65675, 

66853, 66944, 67213 (7). 

Lundell, C. L. 6307 (26). 

MacElwee, A. 2072 (7). 

Mackenzie, K. K. 4764 (7). 

Macoun, — . 361 (7). 

Maire, E. E. 14 (16); 22 (6); 81 (18); 

702 (16); 254 (18); 385, 393 (6); 

1986, 3633 (18); 39 1 4, 6224 ( 6 ) ; 7256 

Marie- Victorin 1 6289, 1 853 4 (7). 
Marie-Victorin & Rolland-Germain 45524, 

49811 (7). 
Marie-Victorin, Rolland-Germain & M. Ray- 
mond 56097 (7). 
Maries, C. 22 (la). 
Martens, J. W. 65 (7). 
Matuda, E. 0529, 5148, 16369 (29). 
McClure, F. A. 3544, 7938 (10); 97<?2 

McClure, F. A. & Chan Kwai Shang 4 

McCoy, S. 3805 (7). 
McCree, H., Jr. 711 (7). 
McGregor, R. C. 1 1 27 5 (lb). 
McGregor, R. L. 645 (7). 
McVaugh, R. 220, 605, 759 (7); 10246, 

IO308, 13906 (28). 
Meebold, A. 8055 (la); 15912 (22a). 
Merrill, E. D. 944 (lb); 9311 (lc); 

IOO70, I02 I 2 (10). 
Merrill, E. D. et al. 15640 (29). 
Merrill, G. M. 323, IOI9 (7). 
Metcalf, F. P. 2030, 2082 (7). 
Metcalf, F. P. & J. C. Chang 605, 718 

(10); 723 (19). 
Metz, — . 1549 (la). 
Meyer, F. N. 1722 (6). 
Minshall, W. H. 1 29, 300, 3590 (7). 


[Vol. 42 


Mizushima, M. I Q2Q (17). 
Mokim, Shaik 266 (lc). 
Moldenke, H. N. 162 3 , 2852 (7). 
Montgomery, F. H. 649 (7). 
Moore, H. E. 2968 (29). 
Morris, E. L. 7775, 1432 (7). 
Morrison, J. L. nig (7). 
Mousset, J. P. 7777 (lb). 
Moyle, J. B. 41 1 (7). 
Muenscher, W. C. & A. R. Bechtel 


Muenscher, W. C. & R. T. Clausen 4680 

Mukherjee, Revoti Mohon 32 (la). 
Mumbauer, J. R. 359 (7). 

Nakamura, T. 36 1 8 (19). 
Narayanaswami, V. 10 5 (la). 
Nath, Amar 24 (la). 
Norton, J. B. <5<9 a (7); 7J4<5 (20). 

Ohwi, J. 436 (17). 
Ohwi, J. & Okamoto, K. 
Oldham, R. 

5°4> 506 
161 (19); 162 (17). 
Ome, Tsuro (?) 


Otanes, F. 

500 (19). 
17985 (lb). 

Ownbey, M. 7570 (7). 

Painter, J. H. 7794 (7). 

Palmer, E. 706 (28). 

Palmer, E. J. 772, 772, 13993, 15237* 
16145, 18923, 21238, 22124, 22814, 
26057, 26960, 27583, 28521, 28785, 

3<?i7> 3JI34, 32933> 34170, 35954, 
30053, 30900, 37i3i> 41926-A, 43587, 
45308, 45570, 45645, 490i6 (7) 

Pammel, L. H. 43, 269 (7). 

Pammel, L. H., H. H. Hume & S. R. Fitz 

1647 (7). 
Pammel, L. H., H. E. Pammel & P. S. Mo 

Nutt 522 (7). 
Paradhan, K. S. 34 (la). 
Parham, H. B. R. (Mrs.) 3 (4). 
Pease, A. S. 14, 3410, 3671, 11963 (7). 
Peattie, D. C. 48 A (7). 
Pennell, F. W. 1 1 48, 16382, 16487 (7). 
Petelot, A. 340, 863, 991 (10); 3184 

(8);5&p (22b) 55937 (8) ; 5936 (21); 

5941 (11); 5946 (22b); 6368 (8). 

Peters, J. E. 29 (7). 

Phelps, O. P. 647, 648 (7). 

Pierre, L. 895 (la); 2794 (lc). 

Poilane, — . 287 1 4 (10). 

Poole, S. F. 102 (7). 

Pradham, — . 34 (la). 

Dr. Prain's collector 70, 78 (la); 289 

(12); 4jp (22a); 795, W02 (8). 
Pratt, — . (55 (7). 

Prazer, J. C. 54, 167 (la) ; 2?<5 (lc) ; 273 

Pretz, H. W. 77403 (7). 
Pringle, C. G. 6842, 6998 (28); £733 

Purdom, W. 944, 945 (17); 946, 946* 

Purpus, C. A. 7094 (26); 7370 (25); 
8080, 8926 (26). 

Ram, B. 309 (la). 
Ramaswami, M. S. 1 490 (la). 
Ramos, M. & Deroy, D. 22511 (lb). 
Ramos, M. & G. Edano 38961 (9). 
Ramos, R. 9, 346, 1813, 42711 (lb). 
Randolph, L. F. & F. R. Randolph 35 (7). 
Rao, M. R. 75<?£ (la). 
Raup, H. M. 7486, 7568 (7). 
Redfleld, I. H. 106 1 (7). 
Reko, B. P. 4019 (29). 
Richards, C. D. 2453 (7). 
Ridley, H. N. 404, 5237, 7127, 8228, 

13538 (9). 
Robinson, B. L. 550, 798, 831 (7). 
Robinson, C. B. 11617 (lb). 
Robbins, G. T. 2292 (7). 
Rock, J. F. 77, 1024 (8); 1744 (10) 

2237, 2770, 2823, 2900 ( 8 ) ; 3556 (11) 

J7PP (^);3885(is); 3931, 4032(11) 

4195,6809 (16); 6946 (18); 7568 (16) 
<?3^ (18); £540 (11); 24679 (16). 

Rodger, A. 198 (lc). 

Rodriguez, G. 2025 (la). 

Rogers, C. Gilbert 864 (lc). 

Rose, D. K. 

3 (7). 
Rosendahl, C. O. 

Rosthorn, A. v. 

3*5 (7). 

9 (10); 1556b (22b); 
7572, 7573 (18). 
Rydberg, P. A. 38, 1453 (7). 
Rydberg, P. A. & — . Ralphlmer 

Ruth, A. 82, 356,369 (7). 


Samnasena, N. S. 34 (la). 

Santapau, H. 87.6, 87.7, 87.8, 87.9, 515, 

570, 603, 1229, 1235, 1967, 1968, 1975, 

2W0, 4102, 8710, 12707, 12708, 14258 

Sawada, T. 2161 (17). 

Schaeffer, R L., Jr. 3503, 588 1, 772 1, 
8770, 8776, 8777, 9821, 9979, 14728, 
I5IOO, 20521, 25023, 29418, 30116, 

31768, 32768, 33851, 37752, 40474 
\' ) • 

Schneider, C. 644, 646, 647, 1027 (11); 

"A 1 * I375 (18); T ^4, 2839 (11); 
2683 (18); 2874, 4112, 4130 (11). 

Seymour, F. C. 1268 (7). 



Shanks, R. E., A. J. Sharp & E. Clebach Tai, F. C. et al. 5142 (11). 


5058 (7). 
Shannon, J. I. 134 (7). 

Sharp, A. J., E. Clebach & A. Clebach 
6309 (7). 

Sharp, A. J., S. Fairchild & E. Clebach 

8254 (7). 
Sherflf, E. E. 1604 (7). 

Shiota, K. 3899, 3905, 9643, 9766 (17). 

Silvestri, C. 1 3 36, 5099 (6). 

Sin, S. S. 11835 (16); 11889 (23). 

Sinclair, J. 39284 (9). 

Singh, A. 25 (la). 

Singh, U. 43 (la). 

Smith, — & — Cave 964 (22a). 

Smith, A. C. 1 122, 4301, 3827, 6085 (4). 

Smith, H. 740, 959, 3743, 6154, 7697 

Tai, L. Y. 204, 107 1 (11); 1 1 13, 1 1 20 

Tang, C. C. & S. E. Ma 2936 (23). 

Tang, S. G. 1677 (19) J 6991, 7039 ( 10) ; 

7085, 13657 (19); 15609 (23). 
Tang, T. 889 (17). 


4674 (7). 

Taquet, T. 632, 2723 (24). 

Tatnall, R. R. 2098, 3633, 4770, 4961, 

4992 (7). 
Taylor, T. M. C. et al. 1239 (7). 
Ten, S. 40, 358, 550 (11). 
Teng, S. W. 90254 (16); 90315 (6). 
Teysmann, J. E. 8731 (lb). 
Thomson, — . 740 (la). 


'898 (16); 13326 (11); 13448 Thone, RH." 214 (7). 

Smith, H. H. 7142, 7785 (7) . 

Smith, U. C. 244 (7). 

Soper, J. H. 2457, 2599 (7). 

Sowett, H. B. S. 752 (lc). 

Spire, — . 289 (lc). 

Standley, P. C. 40524 (7); 71349 (26). 

Standley, P. C. & J. Valerio 52024 (26). 

Steenis, C. G. G. J. van 8717, 9651 (9). 

Stevens, G. W. 1351, 1824, 4173 (7). 

Steward, A. N. 1234, 5470 (16). 

Steward, A. N. & H. C. Cheo 87, 796 
(22b); 429 (18). 

Steward, A. N. et al. 97, 541 (6). 

Stewart, C. C. 1974 (7). 

Stewart, R. R. ////* (i a ). 

Steyermark, J. A. 887, 7572, 12131, 
12875, I3583> 20596, 20825, 22955, 
23533, 24382, 24489, 24563, 24679, 
24874, 25065, 26348 (7); 57015 (26). 

Stocks, J. E. 231 (la). 

Stone, W. 3821, 3921, 10200, 10682, 
10847, I 3319, 13375 (7). 

Strachey, R. & J. E. Winterbottom 2 (la). 

Strathdickie, L. 1 124 (3). 

Stresemann, E. 90 (9). 

Thwaites, C. P. 1232 (la). 
To, K. P. 1868 (10). 

To, K. P. et al. 

Wong 2784 (20). 

631 (23); 889 (10). 


41 88 1 (7). 

Sulit, M. D. 

9917, 10086 

Sun, C. L. 781, 1247 (16). 
Sun, S. C. 1189 (22b); /o* (17); 1189 

Suzuki, S. 1 1 162 (10); 4440.20, 455009, Tsang, W. T. 

Toppin, S. M. 4292, 6154 (10). 
Toroes, R. S. 4009 (9). 
Tso, C. L. 21 1 52 (6). 
Travis, M. T. 2871 (7). 
Trelease, W. 62, 63 (7). 
True, R. H. 92, 11 48 (7). 

True, R. H. & J. K. Edwards 5396 (7). 

Tsai, H. T. 51340, 51343 (18); 51546, 

5 I 591, 51730 (8); 51945 (21); 52375 

(10); 52477 (8); 52545 (10); 52552 
(8); 5290/ (18); 54026 (11); 54565, 

545/tf (22b); 54799 (21); 54949 (22b); 
55076, 55^39 ( 8 ) ! 5542/ (21); 55670 
(8); 5dopd (18); 56542 (10); 569/7 

( 8 ) ; 57255, 5727/, 57545, 575' o, 57769 

(18); 57*99 (11); 57974 (18); 5*555 
(11); 595^5 (16); 59548 (18); 59709 

(16); 59824 (11); 60/52 (8); 6o2<5/ 
(21); 60420 (22b); 60459 (8); 60499 
(21); 607c?/ (18); 609/9 (22b); 60933 
(18); d/004 (11); 67/56, 61629 (22b) 
67(5*0 (8); 61753 (18); 62295 (8) 
d259<5 (21); 62841 (22b); 65059 (18) 
631 16 (16). 


47oo/?> 4^024, 493027 (17). 

Suzuki, T. 6/95 (20). 
Swartley, J. C. 

Taam, Y. W. 578 (10); 720 (8); 724 
(23); 75P (10); /04/ (23); 1477 (10); 
'792 (8); 1798, 1951 (10); 20*7 (8); 
2147 (10). 

Tai, F. C. & C. M. Feng 525* (22b). 

77, 16566 (10); 20092 
(23); 20209, 20604 (1°); 20<575 (6); 

20*55 (16); 2//<?9, 2/4/6 (23); 21422 

(10); 2/75/, 22079 (8); 23574 (23); 
24/06 (8); 24131 (22b); 24/**, 24755 
(8); 23019 (10); 25/^9 (16); 2526* 

(23); 26404 

27666 (6); 2*565 

(22b) ; 28843 (23); 2*927 (20); 38365 

ane. W. 

Fung 241 (10). 


[Vol. 42 


Tsang, W. T. & K. C. Wong 2661 (10); Wang, W. H. 8096 (10). 

208l (8); 2928 (20). 
Tsiang, Y. 720 (8); 1002 (10); 1 346 

Wang, Y. K. 3056 (8) ; 3060 (23) ; 3106 


(16); 1393 (20); 74/9 (16); J644, 

1637 (10); 1705 (8)5 'PJ9> 2300 (10); 
466/ (16); 4945, 4.970 (6); 50/5, 5360 

(16) ; 5543, 5601, 5614(1%); 6450 (16) ; 

6832 (22b); 7499(18); ^J (6) ; 9354, 

95/3 (18); /OO47 (16); /0460 (23). 

C. L. 20788, 20796, 21026 



Ward, F. K. 9362 (21); 9455 (12). 
Waterfall, U. T. 1715, 2830 (7). 
Watkins, I. 5 J (7). 
Watson, J. 176 (la). 

White, C. T. 1943, /2J/7 ( 3 )- 
Whitford, H. N. 81, 414 (lb). 
Whitney, E. G. 

2/132 (6); 21640 (23). 

Wiegand, K. M. 

Tso, C. L. & Y. Tsiang 2040 (10). 
Tsoong, — . 82941 (22b). 
Tsui, T. M. /3#(23); 34? (8) ; 39/ (10) ; 
5^2 (18); 763 (16). 

231 1 (7). 

2765. <5794> 10248 ( 7 ) . 
Williams, L. O. & A. Molina R. 
Wilson, E. H. 


431 (7). 

Uno, K. 21823 ( 24 ) . 

Valeton, T. 102 (10). 

Versteegh, — . /<5^3 

/O (29). 
/<?/ (17);337 (ii); 357* 

(18); 362 (15); 363 (16) ; 364, 364a, b 

(6); 430 (10); 467 (6); 302 (16); 

303, 560 (18); 3d/, 36/ A (10); 688, 

706 (18); 7<*>£ (11); 753, 753A (6); 

P'9 (16); 952, 952 A, 972 (11); /063 

(15); I 106, II 48 

1176 (14); U84 (22b); 



72o > 9 


Vicary (?), 
Visher, S. S. 



2110 (7). 

Voogd, C. N. H. de 682, 2338 (lb) 

Wadmond, S. C. 425 (7). 

Wagner,—. 68 1 6 (7). 

Wahl, H. A. 4959 (7). 

Walker, E. H. 57 1 7 (17). 

Wallich, N. 431 1 (8). 

Walter, N. & C. M. Bangham 

Wang, C. 32148 ( 1 8 ) ; 33029 ( 1 c ) ; 35222, 

35893, 36064, 36607, 40415 (8). 
Wang, C. W. 62839 (11); 62990, 63709 

(16); 64147, 68293 (18); 68718 (la); 

69361, 71910 (18); 72857 (8); 72996 

( la) ; 734/3, 73379, 73390, 73825, 73839, 

870 (10). 


1302, I 519, 2305, 2 305 A (16); 2306 
(15); 2307(10); 230o'(17); 2309 (18); 
23/O (11); 231 1 (18); 23/2 ( 14) ; 2215a 

(16); 3J24 (10); 3325, 41 i 7, 4122, 

4/37(H); 4187, 4195(H); 4317(H); 
478l (18); #£2(11); 6121,6304(19); 
6627, 6907, 7412(17); 8300(10); 8450 
(24); 86ll, 8801, 9203 (17); 9981 
(20); 10425 (24); 10461 (17); //09<5 
(19); II214 (20). 
Winckel, W. F. 


92P, 98P, 1633P, 1682? 

Wood, C. E., Jr. 2710, 3383, 3784 (7). 

Wray, L., Jr. 
Wright, C. 

175, 103 1 (9). 

, 9 (8); 54 (19); 91 (10); 

92 (8); 6/7 (10). 


(8); 74570 
(8); 74966 

(la); 74630 (la); Yamada, K. 


74907 (8); 749OO (lc); 75024 

75183 (8); 7555/ (la); 75846 (10); 

73920 (8); 76355, 76644 (la); 77202* 
(8); 77541, 77833 (lc); 7<?436', 790/3, 

79459 (8); 79407, 7947<> (10); 79560 

(lc); 7977^ (8); 0*0/36, 8018 1 

80194 (lc); <J0232 (10); 80312 

80588, 80781 (8); o"/^ (10). 
Wang, C. W. & Y. Liu 82917 (8). 
Wang, F. T. 21 1 24 (11); 21670 (6); 

22451 (16); 22462 (11); 22633 (6); 

22643 (11)- 
Wang, K. M. 3191 (10). 
Wang, T. H. 


1215 (20). 
Yamamoto, Y. 802 (19). 
Yamazala, T. 44, 44- A (17). 
York, H. H. 733 (7). 
Yu, T. T. 445 (22b); 750, 940, 951 ( 1 1 ) ; 

I7J6 (18) ; 5522 (11); 5562, 577I (16) ; 
5845, 7204, 7978, 8185 (18); 8257, 
1 1 228, 1 1200 (11); 1 1 390 (18); 7/333 
(11); 14375 (lb); 14848 (18); 
(lc); /630* (8); 16592 (22b); 




17294 (21); /7495 (8); 17893 
(18); 17952, 21012 (21). 

Yuncker, T. G. & W. H. Welch 

Wang, T. P. 

7/4 (18); 299, 399 (16). 
776 (17). 

Zinck, N. H. 740 (7). 






Roman type indicates accepted, preexisting names; italics indicates synonyms; bold 
face indicates new names or combinations. 

Acanthothamnus aphyllus 285 

Alsodeia glabra 231 

Catha benthamii 244 

monosperma 249 

vitiensis 291 

Ceanothus paniculatus 230 

Celastrus 227 

aculeatus 274 

acuminatum 285 

adenophylla 285 

alatus 285 

alpestris 285 

angulatus 239 

annamensis 244 

aphyllus 285 

apoensis 246 

approximata _ 250 

aquifolius 285 


var. cuneata 

var. orbiculata 


var. pubescens 

var. papillosus 

var. punctatus 

var. stephanotiifolia 









bullatus _. 




cavaleriei _ .265, 

championii _„_ 244, 


chungii ____ _ 





confertus __ 286 

crassifolia 273 






_ 266 








„ 285 

..- 285 

„- - 240 

...-. 285 

~~ 249 

..-. 282 



... 282 

... 286 

... 276 

... 286 

... 280 

- 276 

crenatus „__ 
crenulatus _ 



.. 286 
— 286 


_.... 286 

cunninghamii 286 

var. parvi flora 286 

dependens 231 

dilatatus 286 

discolor 28 5 

disperma 286 

diversi folia 286 

dubia 286 

elevativenus 267 

emarginatus 286 

esquirolianus 286 

esquirolii ... 287 

euonymoidea 287 

euphlebiphyllus 231 

fasciculatus 287 

feddei 287 

finlaysonianus ... .„ 287 

flagellaris 276 

floribundus 287 

fournieri 287 

geminiflorus 267 

gemmata _„ 258 

gemmatus .... __ 258 

glaucophyllus 253 

glaucus ... __ 287 

gracillimus 267 

grenadensis 285 

haenkea „ _ . 287 

hamelii 287 

heterophyllus 287 

heyneana 287 

hindsii 249 

var. henry i __ 249 

hirsutus 270 

hookeri 254 

hookeri _ 274 

hypoglaucus 257 

hypoleucus 257 

f. P. argutior 257 

f. a. genuina 257 

f. 7. puberula _ _„„ 273 

ilicifolius _ 287 

insularis „ _ 261 

japonicus _ _. 287 

jeholensis _ _. 261 

jodinii 287 

kiusianus 266 

kouytchensis 287 

kusanoi - 



- 268 
_ 261 
... 231 


[Vol. 42 


leiocarpus . 



linearis var. madagascariensis 




lokcbongensis __. 



lucid a ... 





- 279 







2 8 1 



lycioides 288 



magellanicus .... 








merrillii _ 
metzianus ... 
mexicanus ... 
micrantha „ 

.. 288 
. 288 
_ 238 
„ 288 
_ 279 
. 288 
_ 246 
- 250 

- 288 
_ 255 
.. 284 
„ 250 
_ 230 

- 288 

- 288 



monospermus .. 



multiflorus ... 

... 285 

... 288 

- 249 
„ 246 

- 244 

- 288 
.- 288 

- 288 

myrtifolius 288 

neglect a 272 , 



oblanceifolia _ 
oblongifolius _ 







obtusifolia 288 


opposita ... 


f. tnaior 


- 288 

- 289 

- 289 

micropbylla ____ 
var. aureo-arillata 



f. papillosus .. 
var. papillosus 
var. pilosus ___ 
var. punctatus 











oxyphyllus 289 

pachyrachis 279 

pallidus 289 




var. and am an ic a 
var. Balansae ... 


multiflorus (Roxb.) Ding Hou 


serratus (Blanco) Ding Hou 




pauciflora 289 


23 1 


pringlei ___ 

pubescens 230 



var. microphyllus 








- 289 

-- 279 

28 5 















rothianus 290 

royleanus _ 290 


- _ „ 253 






var. trinitensis __ 







7. latifolius 

5. cochlearifolius 
e. elaeodendroides 



robustus „ 



rugosus _ 
scandens __ 
scandens __ 





sepiarius 290 

serrulatus 290 

siltepecanus 282 

spicatus 290 

spiciformis 256 

var. laevis 256 

spinif olius 290 

spinosus 290 

stephanotiifolius 261 

striatum 266, 290 

strigillosus 261 

stylosa 254, 265 

var. Loeseneri 254 


ssp. stylosus 272 

ssp. glaber 273 

suaveolens 290 

subspicatus 236 

tatarinowii 260 

tetramerus 290 

tristis 290 

uncinatus 290 

vanioti 256 

variabilis 290 

venulosus 249 

versicolor 261 

verticillatus 290 

verticillata 290 

vitiensis 291 

vulcanicolus ____ 282 

wallichianus 291 

wallichii 291 

wightianus 29 1 

yunnanensis 291 

zeylanica 291 

Chailletia gelonioides 285 

Colubrina ferruginosa 289 

Diosma serrata 231 

Elaeodendron glaucum 287 

Embelia esquirolii 258 

Erythrospermum hypoleucum 257 

Euonymoides 227 

scandens 240 


altus 290 

cuneif olius 286 

euphlebiphyllus 231 

thunbergianus 285 

Flueggea serrata 249 

Gouania corylifolia 290 


f eddei 287 

henryi _ 287 


acuminata 289 

crenata 286 

diversif olia 286 

emarginata 286 

heyneana 287 

montana 286, 288, 

neglec ta 2 8 5 , 


neglec ta 272 

ovata 289 







trigyna 288 

variabilis 290 

vitiensis 286, 291 


286, 289, 291 




cuneif olia 287 





scandens 286 

suaveolens 290 

Iodina rhombifolia 289 

Kurrimia sp. 289 



_ 285 

boaria 28 5, 288, 290 

buxif olius 289 


... 289 
... 286 

cunninghamii 286 

dispermus 286 

elaeodendroides 289 

fournieri — 


meridensis _ 



2 8 8 



orbicularis 286, 288, 289 

ovata 289 

quadrangulatus 289 

racemosus 279 

retusa 286, 289 

truncatus 287 

verticillatus 290 

williamsii 282 

Microtropis bivalvis 285 


Moya spinosa 

Myrsine semiserrata — 2 86, 

Orixa japonica _ 286, 287, 

Perrottetia alpestris — 

Phyllanthus sp. 

Pittosporum floribundum 288, 

Pleurostylia cochinchinensis 289, 

Prema parvilimba — — 

Prunus myrtifolia 


.. 291 
- 288 

286, 287 









parviflora var. nitidissima 287 

[Vol. 42, 1955] 


yunnanensis 288 Sempervirentes 242 

Saurauja vaniota 256 Simmondsia californica 288 

Schoepfia flexuosa 287 Trichilia jodinii 287 

Scutia commersonii 291 Tripterygium wilfordii 286 

Scutia (?) paniculata 230 Wimmeria mexicana 288 

Semarilla 227 



Emeritus Director 
George T. Moore 

Edgar Anderson 

Assistant Director 
Hugh C Cutler 

Carroll V. Dodge, 


Robert E. Woodson, Jr 

Senior Taxonomist 

Henry N. Andrews, 


Rolla M. Tryon, 

Assistant Curator of the 

Hugh C Cutler, 

Curator Museum of 
Economic Plants 

Gerald Ulrtci 

Business Manager 

George B. Van Schaack, 

Acting Curator of Herbarium 

Julian A. Steyermark, 

Honorary Research 

Frederick G. Meyer, 


Alice F. Tryon, 

Research Associate 


Research Associate 

Nell C Horner, 

Librarian and Editor 
of Publications 


John S. Lehmann 

Daniel K. Catlin 

Second Vice-President 



Leicester B. Faust 

Dudley French 
Henry Hitchcock 

Richard J. Lockwood 
He \y B. Pflager 

A. Vessel Shafleigh 

Robert Brookings Smith 

Ja F. Morreli 




Arthur G Lichtenberger 

Bishop 01 he Diocese of 


Stratford Lei Morton 

Pre lent of the Academy 
Science c uis 

Ethan A. 1 Shepley, 

Chs ishinj 

Un p 

Raymond R. 7 cker, 

Mayor f the < it, Lou: 

Gerald L r, Secretary 

Volume XLII 

Number 4 


of the 




te Food of a Hindu Village of North 

Reprinted from Bulletin No. 2, Bureau Statistics 

i p Hi a 

and Economic R 

1937) . . 

General Index to Volume 

United Provinces. 

Viall Wiser, Ph.B. t M.Sc. 303-412 

A 1 







3 WW 


r «1 he 
'he A 9f 

879, I 



of the 

Missouri Botanical Garden 

Quarterly Journal 

airline Scientific Contributions from the 


tanical Garden and the Henry Shaw Sc 




with the Missouri Botancal 


The Annals of the Missouri Botanical Garden appears four times 
during the calendar year: February, May, September, and November, Four 
numbers constitute a volume. 





Contents of previous 


the Annals of the Missouri 

Garden are listed in the Agri 


the H. W. Wilso 


of the 

Missouri Botanical Garden 

Vol. 42 NOVEMBER, 1955 No. 4 


By Charlotte Viall Wiser 

[Mrs. Wiser's monograph originally appeared as Bulletin No. 2 of the Bureau of Statistics and 
Economic Research of the United Provinces, published in Allahabad, U. P., India, in 1937. It was 
based on data gathered by Mrs. Wiser and her husband, a distinguished sociologist, when they lived 
in the village of Karimpur at long intervals between 1925 and 1931. She eventually worked it up 
as a master's thesis in the Department of Foods and Nutrition of Cornell University, Ithaca, New 
York. Since copies of it are now almost impossible to obtain, either in India or the United States, 
we have obtained permission to reprint it from Mrs. Wiser, from the provincial authorities, and 
from Cornell University. 

A single copy of it loaned to the Library of the Missouri Botanical Garden has been out-of-town 
on loan during a good part of the time since it was received. As an accurate, detailed, and per- 
ceptive account of what plants were grown for which purposes in a North Indian village, it is on 
its way to becoming a classic in the field of Ethno-botany. — E. Anderson.] 

As an aid to bibliographers the original title-pages are here cited in full. Throughout the 
paper a determined effort has been made to reproduce the text exactly, even to the variant spelling. 
Nothing has been changed from the original copy except the references to page numbers. 

The half-title is as follows: 



Foods of a Hindu Village 



The title-page bears the following: 

Bureau of Statistics and Economic Research, United Provinces 


Foods of a Hindu Village 





July, 1936 


Superintendent, Printing and Stationery, United Provinces 




I — Description of the village of Karimpur and its people 

II — Food Sources 

• • 

A — Variety of foods made possible by climate and soil 

B — The crops . . 

(1) Cereals . . 

(2) Pulses or legumes 

(3) Vegetables . . 

(4) Fruits . . 

C — Animal products . . 

D — Spices . . 

E — Food costs in Karimpur 

III — From field to serving tray 

A — Care of foods at harvest 
B — Storage and preservation 
C — Equipment . . 

D — Dishes prepared from 

• • 

* • 


( 1 ) Cereals 

(2) Pulses 

( 3 ) Vegetables 

(4) "Greens 

(5) Fruits . . 

(6) Animal products 
E — Festival dishes . . 

IV — Food Practices . . 

A — Taboos . . 

B — Nutrition practices 
V — Value of foods now in use 

A — Jail dietaries of the Province 
B — The work of Mr.r.AiisKn\i 

C — Foods available in 


VI — Conclusion . . 

A — Recommendations . . 

B — The Problem (Summary) 

Karimpur families 














We moved to the village of Karimpur in the autumn of 1925. My husband 
and I had found during our years in India that if we were to understand the people 
of the villages, we must not be content to visit them, but we must live among 
them. After a study of the villages in the district, we chose Karimpur for our 
village home. Uninvited, we moved our tents and belongings into a mango grove 
on the edge of the village. In our ignorance we thought that our village neigh- 
bours would be glad to see us. But they showed no inclination to welcome white 
strangers, and were decidedly aloof until we had proven ourselves harmless. When 
they did finally accept us, they took our residence among them seriously, and saw 
no reason why we should ever leave. 

We thought, in the beginning, that a few months would give us what we ought 
to know of custom, activity, and thought. But when the hot season of 1926 
arrived and we were obliged to retreat to a bungalow in the town of Mainpuri, five 
miles away, we had just begun to appreciate our ignorance. We commuted to the 
village from the town during the summer, and returned to live in it as soon as 
the heat was bearable in tents. We repeated this migration annually until the 
spring of 1931, when we left for America. 

My intention was that I should act as the good will of the study. With one 
son five, and another not yet two, I felt that I had all that I could undertake. But 
the village women, confined to their courtyards, could not come to me. They 
kept calling me to them. And as I saw more of them and their share in family 
activities, I realized that to understand and interpret the life of the village, we 
must know more about the women. They are of much more importance, eco- 
nomically, than most students of Indian villages have given them credit for being. 
Observations in this field led me into further interest in their share in the care and 
preparation of foods. 

Meanwhile, the increasing number of calls on us for medical help impressed 
on us the futility of trying to cure ills while the causes of those ills went unques- 
tioned. This led me into a study of village food from the point of view of health. 
I knew very little about food. But I had to acknowledge that if I wanted to 
understand and help Indian village women I must learn more about food, par- 
ticularly their food. And out of this need grew the present study. 

I learned all that I could about the foods procurable and the foods used, of food 
preparation and food customs. This material I carried to America, and with the 
help of specialists in the fields of nutrition and biological chemistry, I attempted 
to evaluate the foods eaten in our village, and to study the effect of customary 
methods of preparation on food values. The work was supervised by Miss Helen 
Monsch, head of the Department of Foods and Nutrition at Cornell University. 



[Vol. 42 


Miss Monsch and other members of the staff, especially Miss Mary Henry and Miss 
Marion Pfund, helped in the study of the chemistry of foods, and in securing 

standard tables of food values. 


Sumner, Professor of Biological Chemistry, and Dr. L. A. Maynard and Dr. C. M. 
McCay, both of whom have much to bring to the field of human nutrition from 
their researches in animal nutrition. 


Karimpur, the village of our study, is in the northern part of India, in the much 
fought over Ganges plain. This section of the country is known, under British 
rule, as the United Provinces of Agra and Oudh. The latitude of Karimpur is 
27° 9" North, about that of Palm Beach, Florida. The climate, however, is more 
continental than that of Palm Beach, and might better be compared with the same 
latitude in Mexico. The winters are delightful. In October, the heat relaxes, and 
the weather is progressively cooler until January when the mean temperature is 
48.7° F. For us these are the months of refreshment. But for village families 
with lives regulated for hot seasons and no provision for cold, January and Feb- 
ruary are months of misery. In March the hot winds begin to blow, and from then 


F. The 

July: over 60 per cent of the annual rainfall comes in J 

and August. The temperature drops only slightly but the rain brings relief to 
men and fields. September, at the end of the rains, with its heat and humidity, 
unrelieved by rain, is the most trying month for all of us. It is made endurable by 
the thought that the siege of heat cannot last much longer. 

Karimpur is in an area which has played an important part in the history of 
India. Not far away are Kanauj, capital of the Harsha dynasty (a.d. 604) , and 
Kampiyala, the stronghold of the Panchalas, heroes of the Indian epic, Mahabharata 
(before 400 B.C.). Sixty-five miles to the south-west is Agra, with its Taj Mahal, 
and great fort of the Moghuls. 

Agra is the nearest city. Cawnpore, next nearest, is 110 miles south. As yet, 
neither city has made its influence felt in Karimpur. There are two market towns 
which link the village with the outside world. One is Kuraoli, 7 */ 2 miles to the 
north, with a population of 4,717. The other is Mainpuri, 5/ 2 miles south. The 
population of Mainpuri is 15,599. Farmers of Karimpur take their surplus produce 
to either of these market towns to sell, and while on such an expedition, they may 
stop to buy clothing or food not available in the village. But they fear the clever 
merchants of the town, and avoid dealing with them as far as possible. They 
prefer to confine their activities to the village, with the minimum of outside con- 
tacts. Mainpuri is the official headquarters of the district. This necessitates trips 
to the Mainpuri courts to settle major disputes, chiefly those over land. But such 
trips are costly and confusing to farmers. And they prefer, if possible, to settle 
their disagreements at home with the advice of their own village council. 



This aloofness has protected the Hindu culture of Karimpur from inroads of 
changing dynasties — an important reason why it is a good field for our study of 
Hindu foods. It shares this advantage with other villages of the district- But it 
has additional points in its favour. Mainpuri District covers 1,675 square miles 
and includes 1,388 inhabited towns and villages. Two hundred and five of these 
have a populaton of over 1,000, and are too large to be representative. Of the 
1,183 under 1,000, many are very small and confined to few castes, and often to 
only one or two industries. Karimpur, with its population of 754, presents a 
variety of castes (seventeen), as well as two groups or castes of outcastes, four 
groups of Muhammadans, and several families of Christians. It is dominated by 
Brahmans, which strengthens its barriers against foreign ideas. Like most of North 
India, farming is its outstanding occupation, with the subsidiary industries needed 
in a self-supporting community. ■ 

The dusty, mango-shaded highway from Mainpuri to Karimpur, winds north- 
ward through a patchwork of small, oddly shaped farm plots separated from one 
another by low mounds of earth or narrow irrigation ditches. Sometimes there is 
an open stretch of uncultivated land which is too alkaline for crops, but which 
puts forth a stubby growth sufficient for meagre grazing. The animals must keep 
moving and searching if they are to keep alive. One is impressed by the leanness 
of the cows, and the thinness of the small boys who herd them. There are no 
houses on the farm plots. But there are hamlets, not much more than a mile apart, 
where the houses of farmers are huddled together. They are at a distance from the 
highway and easily passed unnoticed, with their monotonous gray walls. 

If one were to follow this highway a few miles further north, he would come 
to the old Grand Trunk Road travelled by Kipling's Kim. Five and a half miles 
out from Mainpuri is the village of Karimpur. There are two small sections of the 
village directly on the highway, while the village proper stands back 200 yards to 
the east. 

The first of the roadside sections is Chamar Nagariya, home of leather- workers. 
The leather-workers are counted as outcastes, but not as untouchables. All of the 
leather- workers' families save one, live in this section. It consists of just two 
enclosures side by side, with a high mud wall extending across the front of both of 
them. Along two sides of each enclosure are small rooms, one room to a family. 
The men who live here wear scant loin wrappings, and no shirt except in cold 
weather. They are dark-skinned. Their legs are long and spindling, and their 
knees and ankles bulge conspicuously. Their ribs and shoulder blades stand out, 
and their arms seem disproportionately long, like their legs. Their faces are deeply 
lined. Their teeth are ground down, and dark. The women's bodies are covered 
by long skirts, badly torn, but still so full that they are a protection. Over their 
heads they wear stained, grayish scarves, and some have short vests. One or two 
have smoother lines than the men, but the others have the same long thin arms and 
hard, bony hands. The boys and girls are dressed like their elders, and look like 
them except for their fuller faces, fewer wrinkles and better teeth. The babies are 

[Vol. 42 


plump and laughing. They go unclad when it is warm, and wear short vests when 
it is cold. 

The other section by the roadside is Kachhi ka Nagla, hamlet of vegetable 
growers. They are true farmers by right of caste, along with the family of rice- 
growers which lives in the section with them. Although poor and in debt to 
Brahmans, they are independent, which may account for their living apart from 
the rest of the village. Among them we have the nearest approach to the joint- 
family that we find in the village. High mud walls enclose each joint-family, with 
its animals, its implements and its stores of grain — all of its possessions except land. 
The men, dressed in loin cloths, are short and slight but not as gaunt as the leather- 
workers. Their dark skin fits more smoothly over bone and muscle. One of their 
elders might be mistaken for Gandhi, especially since a cataract operation has made 
it necessary for him to wear spectacles. The women are small and thin, and their 
babies are plump and placid. The older children are copies of their parents. 

As one turns east and bumps over the remnants of irrigation ditches which cut 
across the dirt road, to the village proper, he sees a long line of almost unbroken 

mud walls. 


only by the differences in height and smoothness of different portions of the wall. 
At the north end and at the south of this long wall there is a cluster of small, low- 
walled enclosures, each cut off from contact with the main wall by a lane. These 
are occupied by two castes of outcastes. Those living on the southern end are 
mat-makers, who make their living as farmhands. Their wives serve as midwives. 
The men are small and most of them are as thin as the leather-workers. But the 
arms and shoulders of the young women curve smoothly. And their skirts do not 
hang from protruding hip bones, like the leather-worker women, but from plump 
hips and abdomens. The outcastes at the northern end are the village scavengers, 
and therefore untouchables. They have recently been baptized as Christians, but 
are still regarded as untouchables by the villagers. Here too, the men are thin, 
with three exceptions. The younger women are attractively smooth-skinned, while 
the middle-aged and older women look tired, and their dark skin is drawn tightly 
over broad cheek bones, and hangs loose and dry between sharp elbow and wrist. 
In this group of untouchables one finds the darkest skins, the flattest noses, and 


village — evidences 

caste enclosures. And beyond these enclosures are more like them, with the same 
high walls and small doorways. As far as possible they are built side to side and 
back to back, the only break being an occasional narrow lane or footpath leading 
to enclosures deeper in the village. If one is a stranger, he may follow any of the 
footpaths through the village and see only high gray barriers and blank doorways 
leading to stables. The stable is the entry- way to every enclosure. But if he is a 
friend, he sees the family enclosures overflowing into the paths. Children, puppies 
and goats tumble under his feet. Carpenters and potters carry on their work in 
the lanes before their doorways. And serving women salam him as thev oass bv 



with head scarves drawn down over their faces. No matter how good a friend he 
may be, he never enters an enclosure other than his own. If it is necessary for a 
carpenter to repair a door frame or a beam, a man of the house accompanies him 
inside and remains until the work is finished. Women of serving castes are obliged 
to leave their own homes to work, but those of higher castes are jealously hidden 
within high-walled courtyards behind high-walled stables. 

The families of each caste live close together, preferably in adjoining en- 
closures. The Brahmans (188) occupy the whole eastern section of the village. 
They are of priest-class by birth, but in practice they are farmers. Below them in 
the caste scale, in the order of their rank, are: bards (15), accountants (6), gold- 
smiths (10), flower growers (17), vegetable growers (152), rice growers (6), 
carpenters (42), barbers (2), water bearers (83), shepherds (26), grain parchers 
(10), seamsters (21), potters (9), tradesmen (14), oil pressers (10). In addition 
there are the outastes — washermen (6), mat-makers (28), and leather- workers 
(29). There are also the Christian sweepers (35); and four groups of Muham- 
madans — faquirs or beggars (23), bangle sellers (10), cotton carders (9), and 
dancing girls (3). The numbers given include all members of each caste, not just 
the members who are actually engaged in the caste industry. All of these, caste 


services with each other. 

It is comparatively easy to recognize a Brahman. Even the older men among 
them who are bowed and withered, retain the bearing and features which mark 
them as Brahmans. There are exceptions, but one is fairly safe in picking the 
tallest, straightest, most firm-limbed young men as Brahmans. Their skin is fairer 
and their features finer than those of lower castes. Their noses are high and nar- 
row. The Brahman women, always kept safely in the courtyards of their family 
enclosures, work as hard as other women. But they too have the fair skin, the 
narrow noses and faces, the firm, smooth bodies which set them apart from others 
of the village. None of them is fat. We have known only one fat person in the 
village, a woman of carpenter caste. Her fat was diagnosed as a symptom of her 
diseased condition. Brahman babies seem no plumper than others, and when 
weaned, they pass through the same period of loss in weight and resistance as do low 
caste babies. But if they survive this precarious stage, they surpass other children 
in weight, height, and strength. Smallpox scars and inflamed eyes are less common 
among Brahmans than among other castes. 

Passing down from the Brahmans, the next three castes compare with them 
favourably. The goldsmith brothers are the village wrestlers, and are fed and 
massaged with this in view. Below them there is a decline in the evidences of well- 
being. From appearances it would be impossible to ascertain to what caste a man, 
woman or child of the humbler levels belongs. Sometimes an individual family 
gives its children a little better care than others, but such families are not confined 
to any one caste. A few Muhammadans may be recognized because of their more 
hawk-like noses. And the Christians are set apart by the consciousness oi their 

[Vol. 42 

untouchability and their broader Dravidian features. Aside from these differences, 
all are very much alike. One becomes so accustomed to signs of insufficient food 
and care that his attitude toward standards of living is apt to shift. 

Most of the men and women of the lower castes are angular, with hollows 
where flesh should be smooth. Bodies go unbathed, in contrast to higher castes 
where pour-over baths are supposed to be a part of the day's routine. They sleep 
and work in the same garments, and change them once in two or three weeks, 
whenever the washerman collects them to wash in the village pond. He uses cold 
water, beats the clothes on a rock, and spreads them on the dusty ground to dry. 
After a few washings a garment takes on the shade of gray of the earth and mud 
walls. The men's hair is cut short, with one long lock at the top. The women 
and girls do not use combs, and their hair tangles and mats, and offers refuge to 
multitudes of lice. Smallpox scars and eye infections or disfigurement are so 
common as to excite no comment. Both men and women work to the limit of 
their strength, and are thankful if they can earn enough food to make it possible 

the tasks required of them. They do not hope for more, 
except on festival days when Brahman housewives hand out bread or cakes to the 
women and children of families performing regular services for them. The boys 
who run about the village lanes, or herd cows or sheep are small, lithe, and active, 
ready for any sort of fun. The little girls are the same. But before they are ten, 
they retire demurely to the courtyards where they care for their baby brothers or 
cousins, and imitate their mothers. The life cycle of the women is briefer than 
ours. The girls mature earlier — at ten or eleven—, are married at twelve or 
thirteen, and go to live in their husband's homes when fourteen or fifteen. From 
this time on, they are treated as grown women, capable of carrying a full share of 
the heavy work of farm households. They begin bearing children in the next two 
or three years, and although they do not have many babies, there is a constant sex- 
strain on their bodies. All of their work is done sitting or squatting on the ground, 
which is less wearying than standing. In spite of this, almost every mature woman 
of the village complains of the dragging down of her body. While we, in our early 
thirties, are facing the choice between a career and rearing a family, they are ready 
to sit back and tell their daughters-in-law how to feed their babies. The Indian 
Year Book for 1931 gives the mean age of males as 24.8 and of females as 24.7, 
according to the 1921 census. 



The families of Karimpur retain their customary aloofness in meeting the prob- 
lem of food supply. They prefer to accept the limitations of their own immediate 
environment, rather than venture into strange markets. Salt and spices are brought 
in from the outside, and sold by the small tradesmen. And an occasional purchase 
is made on a grain selling trip, or on a visit to a religious fair. These purchases are 



season or supply within the village. Unfamiliar products have no charm. About 
40 miles from us is a district reputed for its oranges. These oranges are sold cheaply 
in the Mainpuri market during two winter months. But they are not brought 
home to the village. On the other hand, bananas are familiar, because a few are 
grown in the village. And bananas are brought from Mainpuri, when the price 
is sufficiently low. Each year tomatoes are sold in increasing quantities in the 
Mainpuri bazaar, but in the village the tomato is still a strange fruit, of an un- 
desirable colour, and the villagers regard it with suspicion. Their diet is thus 
largely determined by the boundaries of their own fields. And it rises and falls with 
good and lean years. If the rains are favourable, and there have been no hail storms 
or destructive winds or pests, everyone has enough to eat. The villagers have 
learned to count on one or two such years out of every five. If the rains fail, or 
come too late, the whole village cuts rations. If one particular crop is damaged, 
they get along without it until it is harvested the following year. 

Their exclusiveness does not limit their diet as much as one might expect. There 
may not be a variation in foods in one single day or in a week. But over a period 
of months there is considerable variety. The villager accepts his changes in food 
with the changing seasons. He is thus better prepared to welcome each new crop 
as it comes than is the town man who has greater variety each day. Still further, 
although the town man seems to have a greater choice of foods in the bazaar than 
the villager has in his fields, the villager has discovered many edible things which 
vary his diet and which rarely find their way to the town. 


One reason why a greater variety of foods can be produced from a limited area, 
in the United Provinces than in Bengal and further south, is the greater distinction 
between the seasons. We have our hot dry summers when land rests, followed by 
warm rains which make crops of warm climates, like rice and sugarcane, possible. 
The rains are followed by a long dry season, a part of which is much cooler than 
Bengal, and which permits the growing of cold weather crops, such as wheat. One 
can better appreciate the variety of foods in Karimpur, by comparing them with 
the diet of less favoured districts. In his study of the use of food-grains in India, 
Church (1) found that "There are many districts in India where rice forms not 
merely the chief food-stuff but three-fourths or even four-fifths of its total amount. 
In some places it even rises to seven-eighths or to fifteen-sixteenths of the whole 
quantity, as in Burdwan, Dinajpur, Maldah, Kuch Behar, Manbhum, and Darrang; 
other districts might be named in which it constitutes the only food staple." 

Another factor which contributes to the variety of the foods of our village is 
the variety of soils. The soils of Karimpur are typical of those of village com- 
munities of the Ganges plain. The following division of the revenue village is 
given in the latest Settlement report. These reports are made every twenty or 
thirty years by government settlement officers. 

(1) See References. 


[Vol. 42 


Revenue village of Karimpur 

Village sites 
Covered with water 
Otherwise barren 
Culturable land 
Miscellaneous land 

• • 





The 2,103 culturable acres are distributed as follows 

Type of 

Manjha wet 
Manjha dry 
Barha Dumat wet 
Barha Dumat dry 
Barha Bhur wet 
Barha Bhur dry 
Tarai . . 


• * 

• • 




"Gauhan" is the land immediately encircling the village. It is the best 
manured, and is rented at the highest rate. It may be any kind of soil, usually a 
loam, improved by constant manuring. Farmers prefer to grow maize, sugarcane, 
peas and potatoes on this soil, or on the next grade, which is listed as wet and dry 
"Manjha." "Manjha" is the same type of land as "gauhan," but it is the strip just 
outside the "gauhan," and therefore less manured and less desirable. Most of the 
vegetables are grown on one of these two soils, partly because of the improvement 
of the soil, and partly because vegetables require constant care and constant water- 
ing except during the rainy season. The outlying land is known as "barha." 
"Barha dumat" is a mixture of clay and sand in almost equal proportions. It is 
"generally of a rich brownish colour, adhesive without tenacity, briable [friable] 
without looseness, slippery and greasy when wet and with a soapy feeling when dry, 
and cutting like a cheese when ploughed wet." "Barha bhur" is outlying land 
which is "loose and sandy, and quite incapable of retaining moisture." (2) 

The "dumat" will bear a large variety of crops, including wheat, barley, several 
varieties of millet, and pulses. Whereas the "bhur" is so sandy that crops which 
survive with the least irrigation are reserved for it. It is usually limited to spiked 
millet and coarse pulses. "Tarai" is the low lying land, which "includes all classes 
of natural soils from very heavy clays to loose sand." ( 3 ) In our area it is chiefly 

heavy clay. 

On the "tarai," rice is usually grown, and sometimes water melons. In ordi- 
nary years it is swamped during the rainy season. "Maiyar" is "a stiff, unyielding 
clay of a dark colour, shrinking and cracking in dry weather into a network of 

(2), (3) See References. 




fissures, but expanding when moistened into a sticky clayey mass." (4). It can 
be made to produce some of the less desirable pulses, and millets, and perhaps rice. 

Each farmer wants a share of the better land, and not too much of the poorer 
land. This has led to the partitioning of land into small, queerly shaped plots. A 
man may have as many as sixty or seventy plots scattered over the different kinds 
of soil in every direction from the village. R. Mukerjee writes of these scattered 
plots, "Various are the causes which tend to render a peasant's fields so widely 
scattered. In the village communities there was a deliberate attempt to distribute 
plots among the settlers in different soil areas so as to allot plots of different degrees 
of fertility to each . . . The Indian cultivator's farm tends to fly into fragments 
and grows steadily smaller and less regular. As the population increases, the 
holdings on account of the law of succession come to be unduly fragmented." 
And further on he refers to "the pepper pot distribution of holdings." (5) 

The following table shows the proportions of each type of land included in the 
scattered plots of /, one of the largest land-holders of the village, and compares 
the proportions of his holdings with those of the same kinds of land of the whole 


rentage of each kind of land in Karimpur comp 
percentage of same kind of land held by J, a Bra 

Kind of soil 

Barha Dumat 
Barha Bhur 

• • 

• • 



in revenue 


in ]'s 


in J's 

total holding 








"The hereditary holdings show the closest correlation because they conform 
more closely to the original division of land." (6) 

With such an arrangement, each long-established family has a chance to raise 
a variety of crops, even though its acres are few. Poorer farmers, and those whose 
forefathers came in after the original division was made, may have little or none 
of the best land, and a larger share of undesirable land. But even they manage to 
get a variety of crops, by hard work and a resigned acceptance of inferior products 

as part of their inferior lot. 

W. H. Moreland of the Civil Service discusses the management of farm land 
composed of scattered holdings, in his Agriculture of the United Provinces: 

(4), (5), (6) See References. 

[Vol. 42 


"Now, considering the variety of soils and of possible crops, this looks like a very complicated 
problem, and it is fortunate that the individual cultivator has not to iace it without some 
guidance. The question has been attacked by many generations of cultivators and their accumu- 
lated experience is to be found in the custom of the country which guides the individual on such 
questions as what crops can be safely grown on hot soils, what crops will repay irrigation, what 
is the best means of distributing manure over the holding, in what order should crops be grown, 
and so on. This custom of the country is not an infallible guide to the most profitable utiliza- 
tion of the holding; the best course may never have been tried or may have been discarded 
through some mischance, or changes in the level of prices or as in the demand for particular 
products may make some changes in the custom desirable; but it is usually a safe guide to making 
a living, and the ordinary cultivator is well advised in following it rather than applying his 
limited intellect to working out a solution afresh." 

And further on, he describes the arrangement of crops as actually practised: 

"To see how this custom works out in practice, we may take the case of a holding of the 
ordinary size, say 5 or 6 acres, containing different soils. There may be a field of high sandy 
land; it cannot be irrigated and it dries so quickly that it would be very risky to sow winter 
crops on it: this land will usually have a rainy season crop. But even in the rains it is not very 
productive, for it will suffer from either wet or drought: not very much will be spent on its 
tillage, and it will usually be sown with cheap crops such as bajra (small millet) and moth 
(a pulse) which do fairly well on such land at a small expenditure. There may be another field 
of heavy clay; this will be regularly sown in the rains with rice, the only crop for which it is 
really fitted, while if possible some cheap pulse will be sown with rough tillage after the rice has 
been gathered. The rest of the holding consists, we will suppose, of irrigable loam. One portion 
of it, however, is so low lying that it is always more or less flooded in the rains: this portion 
will probably be sown regularly with a winter crop, which may sometimes be wheat and some- 
times a mixture of wheat or barley with gram or peas. The rest of the land is suitable for either 
season and will be divided between the two in such a way that tillage can be effected in the time 
available. Possibly the cultivator will do this in the following way: taking the field nearest his 
house he will apply his manure to it and sow maize, which will be followed by some winter crop: 
another field he will put under great millet, arhar and urd (two pulses), with some sesamum to 
supply his house with oil, and a border of hemp to give fibre for well-ropes, etc.: the remaining 
land will be left for the winter and after manuring, sown with wheat along with a border of 
linseed or some lines of rapeseed, to supply more oil. In this way the cultivator would get an 
early supply of food from his maize, say about the end of August or September: rice in Septem- 
ber or October; juar, bajra, urd and moth (two millets and two pulses) in November: pigeon 
peas and whatever cheap winter crops he had sown in March and April, when he would also 
have his wheat ready to sell for the rent. 

"This is the holding of a cultivator of no special skill: a better man would get in a field of 
sugarcane or opium, or would manage to save more manure and take two crops in a year off a 
larger area, but the general principles will be the same: to secure sufficient food, with preferably 
an early supply in the rains; to have something to sell; to have a good variety of crops; and to 
arrange so that there may be time to till for all with the single pair of cattle which is all that 
a holding of this size can support.'* (7) 


One of the farming customs of the Ganges plain is the mixing of crops. 
Products intended for sale are kept unmixed. At least the farmer tries to keep 
them unmixed. But he is so accustomed to mixtures, that if barley or pigeon peas 
get into the wheat he sees nothing objectionable in the mixture. It is when his 
product reaches the city market, particularly where foreigners buy, that he is 
accused of adulteration. In his study of "Food-Grains of India/' Church remarks, 
"The value of Indian wheats in European markets is often much lowered by pre- 
ventable impurities. Very frequently they contain other cereal grains, especially 
barley; gram and linseed some times occur in them." (8) And McCay in his 
study of jail dietaries in the United Provinces, writes, "The chemical analyses of 
wheat in use in the different jails show greater variation than any other of the 

(7), (8) See References. 



food materials. This is practically entirely due to the different degrees of con- 


class quality and contained a large proportion of foreign grains." (9) 


cereals. Its presence makes bread made of barley or millet more palatable. And 


The more common mixtures have names, according to the grains they contain, and 
the proportions of each. "Gojai" is a mixture of wheat and barley. "Tirra" is 
wheat, barley and gram in equal parts. "Bejhar" may be equal parts of barley and 
peas, or equal parts of barley, gram and peas. Within the village, the selling or 
loaning of mixed grains can be carried on satisfactorily with the help of these 


There are several advantages to the Karimpur farmer in raising mixed crops. 
A mixed crop may serve as insurance. A crop, such as rice, which flourishes in 
heavy rains, may be planted with a coarse millet, which flourishes in dry weather. 
Then, if the ground is flooded there will be rice, and if there is very little rain the 
rice may be lost, but there will be a millet crop. Sometimes two crops are grown 
together to save land. Millet and the pigeon pea are such a combination. The 
millet grows up more rapidly and the pigeon pea grows more slowly, until the 
millet is cut. As soon as the millet is cut the pigeon pea grows more rapidly, and 
when harvested yields about two-thirds of what it would without the millet. And 
the farmer is one millet crop ahead. In some cases a creeping pulse is grown with 
taller crops, to retain as much of the moisture in the ground as possible. Or a 
deep-rooted plant may be grown with one with shallow spreading roots. The 
most common crop mixture is that of at least one pulse with a cereal. The farmer 
does not know about nitrogen but he and his forefathers have found that the soil 
is benefited by the presence of a pulse. They have also discovered that pulse alone 
is not a satisfactory food. The mixture of pulse and cereal is the happy solution 
in the field and on the brass eating tray. The following record of his own crops, 
given us by a village cultivator, reveals the extent of crop mixing, particularly 
with food crops: 

The grain crops of the village are familiar to us of the West except in the use 
for which some of them are intended. We in the United States regard certain 
millets, such as sorghum vulgare, as fodder crops, but in Karimpur they are grown 
as a complete family food — first for the human members of the family, and 
secondly for the animal members. Moreland, in his Agriculture of the United 
Provinces, remarks: "in many parts the cultivator cannot spare any part of his 
land solely for fodder crops, but must grow crops which will feed himself and his 
family as well as his cattle." 

To a Westerner, the pulses are less familiar than the cereals. The "arhar," or 
pigeon pea, is the most common. Its yellow flowers colour the village fields during 
the cold season. It is one of the hardiest of the field crops. The peas, dried and 

(9) See References. 


[Vol. 42 


Crops grown by a Karimptir farmer 




• • 

Field peas . . 
Sarson (mustard) 


Duan (mustard) 



Water melons 
Little millet 

Great millet 

Spiked millet 



Pigeon pea 

Urd (pulse) 

Mung (pulse) 
Moth (pulse) 
Ronsa (pulse) 

Cluster bean 


Grown with 




Barley, or 
Barley and gram 

Gram, or . . 


Alone, or with wheat 

1 . Wheat, or . . 

2. Barley . , 

1. Wheat 

2, Wheat and barley 

1 . Wheat, or . . 

2. Barley . . 



1. Hemp, or . . 

2. Castor plant on border. 


• • 


• • 

1. Pigeon pea, mung, or urd. 

2. Mustard • . 

1. Moth (pulse) or pigeon 

2. Mustard 

• • 


1 . Millet 

2. Cotton 

1 . Millet 

2 . Cotton 


Spiked millet . . 

1. Great millet 

2. Spiked millet 

1. Great millet 

2. Spiked millet 


• • 

Reason for mixing 

1. Mixed food. 

2. Mixed food. 

1. Good for soil, and 

2. Mixed food. 

Mixed food, and good for 

Both save ground. 

Both for mixed food. 

Both save ground. 

Both, 1 and 2, save ground. 

Saves ground. 

1. Mixed food. 

2. Saves ground. 

1. Mixed food. 

2. Saves ground. 





Mixed food. 
Saves ground, 

Mixed food. 
Saves ground 

Mixed food. 
Mixed food. 
Both, mixed food 

Both, mixed food 

Saves ground. 



split, are used by everyone. "Urd" and "mung" are much alike and are usually 
classed together. They are the best liked of all the pulses. "Moth" is the least 
used. Some do not care for it, while others keep a small store of it on hand as a 
special food for invalids. These three, "urd," "mung" and "moth," are dried and 
split, like the pigeon pea. Bengal gram is one of the most welcome of the products 
of the spring harvest. In the fields, around the home fire-place, and at the grain- 
parchers, it supersedes all other parchable foods. Travellers carry it parched, often 
their sole food. In the Indian army it has been adopted as emergency rations. A 
few pulses, like the field pea and the cluster bean, can be used green for a short 
period. But as soon as the peas are ripe they are treated like the other pulses — split 
and dried. When split, this whole group of pulses is known as "dal," an important 
division of any family's diet. 

Vegetable crops are traditionally in the hands of vegetable growers, Kachhis. 
The task of growing them is not so much an exclusive privilege as a laborious 
caste duty. We have been interested in the slight increase in vegetables grown by 
ordinary farmers during our years in the village. They have discovered that enough 
water spills over the edge of the big leather water bags to keep vegetables alive on 
a small patch of ground surrounding an irrigation well. Beyond this they have 
neither the patience nor the interest to struggle with vegetables, especially in the 
hot weather when it seems as though nothing green can possibly survive. They 
leave vegetable growing to the Kachhis. The vine crop is the largest vegetable 
crop in the village. It includes several kinds of cucumbers and melons which were 
new to us. Like other foreigners in India, we found the hot season made almost 
unbearable by the absence of fresh vegetables, until our Indian friends introduced 
us to several unattractive forms of cucumber, melon and gourd, and showed us 
how appeti2ing they could be made with the help of green mangoes and spices. 

Tubers and roots are depended upon for winter vegetables. Potatoes are dug 
and used when very small, chiefly because the crop is in danger of being stolen, or 
of being destroyed by porcupines. Radishes and carrots, on the other hand, are 
allowed to grow very large. In his Agriculture of the United Provinces (p. 222), 
W. H. Moreland reports "Radishes are grown for their bulk, and may be anything 
up to a foot in length." Of carrots he writes, "The Indian carrot is different from 
that which is familiar to Europeans as a vegetable; the root is dark coloured and 
is coarse and flavourless." Beets and tomatoes are grown on a small scale in the 
vicinity of Mainpuri, but Karimpur conservatism bars them from its fields. The 
prejudice is chiefly against their colour which is that of blood. Onions are limited 
in their growing and use to the lower castes. In the Laws of Manu, V. 19, it is 
written "A twice-born man who knowingly eats murshrooms, a village pig, garlic, 
a village cock, onions or leeks, will become an outcaste." 

There appear, in the village list of vegetables, foods which we overlook. Buds 
of certain trees are treated as vegetables. And a whole series of leaves, some wild 
and others incidental to the growing of some field crop, are utilized. Among the 
former is the white goose foot, and among the latter are radish tops, the leaves of 

[Vol. 42 


the Egyptian arum, and the tender tips of pulse and mustard plants. All of these 
leafy foods are set apart from other vegetable foods, and cooked in a special way. 
In the village they are called "sag," or greens. 

The village looks to its trees as a special source of variety in the diet. And the 
trees have something to offer during most of the year. Some of the fruits, like the 
neem and peepal berries, we would never consider as human food, and in govern- 
ment publications they are listed as famine foods. But the children eat them in 
large quantities and farmers munch them on their way to or from the fields. The 
children also gather and eat the blossoms of the mahwa tree, and forage for wild 
plums. Their enjoyment of plums, as well as of other fruits, when hard and 
green, is amazing. Anyone who has tried to eat a tama 
f ullv sour it is. We use it to make a hot season drink. 

inyone who has tried to eat a tamarind pod knows how pain- 
use it to make a hot season drink. But the village youngsters 
chew one after another without a blink. The tamarind is valued in Sanskrit medi- 
cine for its antiscorbutic effect. Certain fruits are so rare that only a few can 
have them, even for a limited season. Among these are the pomegranate, custard 
apple, guava, date and lime. Bananas are somewhat less rare, but are treated as a 
luxury. They are sliced green, skin and all, and cooked with vegetables. Musk- 
melons and water-melons appear when most welcome, in the hot, dry weather. 
Everyone manages to get a share of both. 

The most popular and the most abundant of all the fruits is the mango. In the 
hot months, when the ground is baked hard and few growing things survive, the 
mango trees yield their crop. Mangoes seldom ripen. Wind storms bring down 
some of the fruit. And the rest comes down, with the help of sticks and stones 
and secret trips up the trees, before we think it ready to eat. It is the right of 
everyone in the village to have a share of the mangoes. This is in accord with the 
Laws of Manu (VIII, 339). "The taking of roots and of fruit from trees. . . . 
Manu has declared (to be) no theft." The owners of groves do not object to the 
stripping of their trees, as long as the trees are not damaged, and as long as the 
fruit is reasonably distributed within the village without being sold. If mangoes 
were picked and sold, even by the owner of a grove, the village would deal with 
the offender by its own unique method, that of non-co-operation. We have known 
progressive landlords who have raised mangoes for sale, but they have been power- 
ful enough to assert their rights. In our village, the old custom prevails to the 
extent that owners of groves entertain no thought of financial profit. One man, 
of farmer caste, owns an unfenced grove of 25 old trees, bearing in a good season 
over 50,000 mangoes Our office clerk lived in the home of this farmer, and was 
in a position to observe closely. No mangoes were brought into the owner's house 
from his own grove, during the clerk's two year's stay with him. Fruit from near- 
by trees, belonging to others, met the needs of his family. He was following the 
practice of his forefathers in letting others enjoy the fruit from his trees. And he 
is assured that as they were blessed by the religious merit thus acquired, so will he 
be blessed, especially since a number of those who eat his mangoes are twice-born. 
The Government encourages the preservation of groves by not taxing them as 




long as they remain groves. Rule 4, Chapter I of the Wajib-ul-arz (Customary 
Law) states that, "If the trees of a garden are cut down and new trees are not 
replaced, the Government will have a right to assess land revenue according to 
circle rates, if that grove has been exempted from the assessment of land revenue 
at the time of the current settlement." This has helped retain the groves which 
existed before the last settlement. But there has been almost no effort to plant new 
groves. The custom of free distribution of fruit fitted into the old regime of 
simple, personal exchange. Under the comparatively recent system of money 
exchange, men want some assurance of a money return, before taking the trouble 
to plant and foster young trees. The following table gives a picture of the sources 
of the village supply of cultivated fruits. 

Number and kinds of trees found in groves of Karimpur 

Name of tree 

Mango • . 

Tamarind . . 

Neem . . 

Peepul . . 

Gular (wild fig) 


Wood-apple . . 

Indan jujube 

Jaman (kind of plum) 

Khajur (kind of date) 

Khata labhera 

Heens . . 

Jack fruit . . 

Chuara (kind of date) 


Guava . . 

Plantain . . 
Lime . . 

Mahua . . 

Rai labhera (kind of plum) 




mature trees 




































Number and kind of trees on the village site 

Name of tree 
Mango . . 

Tamarind . . 
Peepul . . 

Ber (Indian jujube) 
Jaman . . 

Khata labhera (plum) 
Anar (pomegranate) 
Amrud (guava) 
Banyan . . 

The ponds near the villa ee contribute a share of the food 









The phapola, puffed, or popped, was oi 
foods. We like it better than popped 

[Vol. 42 


At the festival known as Deotan, in November, everyone in the village is 
allowed a taste of sugarcane, even though it is not ready for cutting. From this 

on. one meets children and men chewine and suckine stalks of cane. A stalk 



Not much fresh cane 

reaches the women shut in the courtyards. But they enjoy whatever the children 
remember to bring home to them. The fresh juice adds variety to a number of 
dishes during the pressing season. And the raw sugar, made by boiling down the 
juice and pressing it down into large cakes, supplies the village with sweetening for 
the year. 


Animals are a minor source of food supply in a vegetarian community like 
Karimpur. Cows are kept and milked, but the milk is seldom used as such. Boiled, 
churned into butter, and clarified, it supplies the pure fat for deep fried foods. 
Beef is never eaten. Goat meat is eaten about twice a month by the goldsmiths. 
And Muhammadans eat it of tener, if they can afford it. The goats kept by the 
village goatherds are supposedly kept for milk, which is sold in Mainpuri. But 
occasionally a goat, no longer useful, is sold quietly to a Muhammadan butcher. 
Any goat meat beyond this comes from Muhammadan communities outside Karim- 

Fish is not eaten by members of the higher castes. But members of lower 
castes have a feast of fish, during the few days before the ponds dry up for the hot 
season. The men who are draining the pond with their reed water-lifting baskets, 
let it be known when the water is getting too low for further usefulness for irriga- 
tion. The word passes swiftly, and on the following morning a crowd of low 
caste men and children gathers at the edge of the pond. Each one has a big round 
basket without a bottom. The instant the irrigators announce that they are 
through, the fishers wade into the shallow water and push their baskets down into 
the mud bottom. They scoop out any fish they have thus captured, and move 
forward a few steps where they try again. The fish are small, but they are an 

annual treat. 


No one else would eat their eggs or their flesh. The sweepers, along with the 


system, and which are the only animals raised with a view to butchering. There 


and is butchered, the feast is theirs. No caste person would come near enough to 


If a family is small, and can afford to keep a buffalo, there will be "ghi" or 
ified butter, for all of the cooking. But in most families the "ghi"* must be 

oil. There are several varieties of 


tard and rape grown in the fields of Karimpur, usually mixed with grain crops, 
harvested separately. Each farmer tries to have his own supply, which he 

"Pronounced "ghee". 




stores, and has pressed month by month, by the oil presser. The better grades are 
very satisfactory for deep-fat frying. The 


The women 

poorer oil, try 

When mustard oil become. , , 

supply. The sesamum is valued not only for its oil but for its flavour. The seeds, 

with crude sug< 
in the autumn. 

These are among the most popular sweets 


If one tries eating Indian vegetables with only salt added, and then visits an 
Indian home where these same vegetables are prepared with spices, he has a new 
appreciation of the art of spicing. An untempting vegetable can be spiced and 
cooked in several ways, and scarcely be recognized as coming from the same plant. 
This ability to use spices properly is even more important where there is no meat 
or meat stock to supply flavour. The variety of spices used in a village home is 
limited by purchasing power. Everyone tries to have at least red and black pepper, 
coriander, asafoetida, and turmeric, along with salt. Other spices are necessary 
for certain dishes, but may be omitted from ordinary daily cooking. A list of 
spices sold by a village tradesman follows. Two other tradesmen have smaller 
stocks of the same things. The list does not include peppermint, which is grown 
in some of the larger courtyards, or coriander, which is grown in fields with grain 

Condiments sold by Mohan Singh, Karimpur 


il name English name 

Botanical name* 


• . Asafoetida . . 

• • 

Ferula alliacea. 

Lai mirch . 

• ♦ Red pepper 

• « 

Capsicum frutescens used both green 

and ripe. 

Kali mirch. 

• . Black pepper 

■ - 

Piper nigrum. From unsorted berries, 
with some husks. 

Safed mirch 

. . White pepper 

• • 

Piper nigrum. From sorted berries. 

Saunf . , 

i • . Aniseed • • 

• • 

Pimpinella anisum. 

Kala jira . , 

, . . Small fennel 

« • 

Nigella sativa. 

Safed jira . 

• • Caraway . . 

• • 

Corum carui. 


, • . Turmeric . . 

• « 

Curcuma longa. 

Laung . , 

» • • Clove . . 

• • 

Eugenia caryophyllata. 

Chhoti elach 

i . ♦ Small cardamom 

• • 

Ellettaria cardamom, true cardamom. 

Bari elachi . , 

, . . Greater cardamom 

• • 

Amomum subulatum. 

Nariyal . , 

. • • Cocoanut . . 

9 • 

Cocos nucifcra. 

Ajwain . « 

• • • • 

Caryota copticum. 

Sonth . « 

, • . 1 Green ginger root 

m * 

Zingiber officinale. 

Kismis . . 

• • 1 Raisins . . 

• • 

Vintis vinifera 

Chiraunji . . 

♦ • • • 

Buchanania latifolia. 

Namak . . 

. . (Salt 

. . • • 

Kala namak 

. . 1 Common salt with sodium carbonate, in sort of fusion. 

Chhoti pipar 

and bari jLong as cloves, but thin. Not quite as hot as pepper. 


— _ 


• . J As thick as cardamom, 
1 digestion. 

and a half inch long. Supposed to aid 

•As in the rest of the paper, the spelling is reproduced here exactly as in the original copy. 



[Vol. 42 



Farmers would 

not have money to buy food at these rates, low as they seem to us. They produce 
most of their own food, or exchange food which they have stored, for food or 
services which others have to offer. When the exchange is made, there is a great 
deal of arguing, and he who is the best barterer gets the best bargain. Either the 
seller or the buyer produces balance scales. One article to be exchanged, such as 


in the other. The exchangers agree on how the scales should hang, according to 
their estimate of the value of each article. The pan of cucumbers should swing 
far down if wheat is in the other pan. If there are sweet potatoes in the other pan, 
they should almost balance the cucumbers. With this in mind, one realizes the 
futility of measuring a farmer's food expenditure in terms of money. At best, it 
can only give a general idea of comparative values. The figures here given we 

be as accurate as can be 

found. They were recorded by our clerk, in the 
household of farmer caste in which he lived. If we were to list the prices which 
we were actually paying for the same articles at the same time, our figures would 
be much higher— twice as high in some cases. A white face always sends prices up. 

Prices of foods in Karimpur in 1928 



Barley . . 
Maize . . 
Rice . . 

Spiked millet 
Great millet 
Gram . . 


Mung . . 

Moth . . 

Pigeon pea 

Ghi . . 

Mustard oil 
Potatoes . . 

Lauka (kind of pumpkin) 
Lauki (kind of gourd) 

Turayan (kind of gourd) 


Egg plant 

Green figs . . 

Bitter melon 

String beans 

Radishes, with tops 

Carrots . . 

Egyptian arum 

Khata labhera (blossoms) 

Khata labhera (green fruit) 

Cluster bean, with 

Cow pea . . 

Field pea . . 


Number of pounds 

purchased for 

1 rupee 



. . 20 









Cucumber . . 

Sweet potatoes . . 

Kacheria (kind of melon) 

Senda (variety of cucumber) 


Mustard tops 

Fenugreek (green) 

White goose foot 
Gram tops . . 
Spinach and dill 
Nori (wild pot-herb) 
Onion tops 
Paintiya . . 

Lissua (wild pot-herb) 
Pea vine tops 
Chaunrai (pot herb) 
Leaves of Egyptian arum 
Mangoes . . 
Guavas . . 

Jaman (kind of plum) 

Water melon 

Ber (Indian jujube) 

Water chestnut 

Tamarind pods 

Musk melon 

Raw sugar 


Red pepper 

Black pepper 

Turmeric . . 

Aniseed . . 

Coriander . . 

Cloves . . 
Big cardamom 
Garlic . . 

Fenugreek ( seeds ) 
Caraway . . 

Number of pounds 
purchased for 








Price of meat 




It was divided into 12 shares plus the head. Each share weighed 
mds, and cost about $.16. The head weighed more, and was the 

most desirable part. Cost $.20. 


[Vol. 42 



At harvest time, all of the able bodied men and boys of the village and some 
of the low caste women go out to the fields. Some cut their own crops, and some 
are employed to cut those of others. The men squat in a row at one end of a field 
and move slowly toward the opposite end, cutting the grain handful by handful 
with their sickles as they hunch along. They tie the cut stalks into bundles, bind- 
ing them with tough grass. When a field is finished, they carry the bundles, usually 
on their heads, to one of the groves beside the village. We camped in one of these 
groves and used to watch the men coming in, balancing their loads and dropping 
them on the large pile belonging to the master of the field. 

The millets are treated differently. Women of farmer and water-carrier castes 
walk slowly through the field breaking off the heads of grain and dropping them 
into large flat baskets which they carry. Later the men cut the stalks. This job 
can be postponed until more important crops are harvested. 

Rice is cut in the usual way, but is not carried to a grove. It is threshed close 
beside the field where it has grown. 

Most of the grains and pulses are threshed by the treading of bullocks on a 
threshing floor. The threshing floor is simply a circle of ground from which grass 
and weeds have been scraped. The threshing-floor for rice is prepared beside the 
rice field. Other threshing-floors are in the groves where the cereal and pulses have 
been stacked. In our grove there were five and sometimes six threshing-floors being 
used. On each of them from two to six bullocks were driven round and round 
day after day. Small boys usually do the driving. A stack of grain about three 
feet high, covering a threshing-floor with a radius of ten feet, takes three days to 
threshed by four bullocks. This varies of course with the activity of the driver. 
People without bullocks must leave their grain stacked in the grove until some one 
else's animals are available. 



In the threshing, the husks are loosened, and man] 
from the kernel. The husks of rice and barley are not 

bullocks. With these two the threshing simply breaks »,v •«*«* AWUI WiW .._. 
The actual husking has to be done later by the women in the courtyards. 

Close after the threshing comes the winnowing. The men hold shallow baskets 
of grain high above their heads, and tilt and jerk them in such a way that the 
kernels fall to the ground. The chaff is carried away by the breeze. If there is 
no air stirring, two men stand behind the winnowers and flap a long sheet, thus 
making a breeze. 


o be sold is weighed out and poured into gunny bags, beside the thresh- 
The two men who keep large store-houses of grain to loan or sell later 



A few others keep their surplus in 




bags. But most of the products of any harvest— grains, pulses, vegetables, or 
fruits— are carried home from threshing-floor, grove or field, and poured out on 
the floor of the courtyard. Whatever remains to he done, ti r h*» wnrk nf *V.- ™™*« 



They must have them made and well dried before the harvest. As rapidly as the 
grain is brought in, they spread it on sheets, on their roped cots, and keep moving 
the cots so that they may get all of the sun possible within the courtyard. When 
the grain is thus thoroughly dried, they clean it basket by basket full, and fill the 
jars with it. Near the base of each jar is a small opening, plugged with rags. 
Through this, a day's supply of grain can be drawn out. 

Pulses to be stored require additional treatment. The women must remove any 
pods not removed in threshing. Then they soak the pulse over night in water and 
oil. The quantity of water varies. To five pounds of pigeon peas they add ten 
pounds of water and three and a half ounces of oil. To five pounds of "mung", 
"urd", or "moth" they add only four ounces of water and an ounce of oil. After 
the soaking they dry the pulse in the sun for a day or a day and a half. For this 
they requisition every cot in the house and every sunny inch of the courtyard. 
When the pulse is thoroughly dried, they split it in the light stone hand mill 
described. If a woman does not own such a mill, she tries to borrow one from a 
neighbour or employer. If she cannot borrow, she pries apart the stones of her 
flour mill, with a wad of cotton on the top of the pivot, and splits the pulse in it. 
She may split her whole season's supply immediately after the harvest, or she may 
take care of enough for three or four months at a time. During the splitting 
process many husks are loosened. She gets rid of them by winnowing the pulse and 
sifting it, before storing it in clay jars. 

There is heavy work left for the women, in the care of rice and barley. Both 

of these have been partially threshed by the treading of bullocks. Bullock's hoofs 

(are able to separate the husks from the kernels of some grains, but not those of 

rice or barley. This task is left for the women. With an iron-bound wooden 

pestle, a woman husks her rice or her barley in a stone mortar set in her floor. It is 

tiring, strenuous work, and she prefers to do a small portion at a time, as it is 

Where there is a supply of a certain grain to be stored, too small to require a 
tall jar, it is stored in smaller jars, made and baked by the village potter. These 
are often found in rope baskets, suspended from beams of the family's store-room, 
pff of the courtyard. Grain kept for seed may be stored in the same way, or hung 
in the courtyard on a beam protruding beyond the store-room wall, protected by 
eaves. The village housewife wages continuous battle with the rats and insects 
which covet her stores, and with the crows which settle on the courtyard wall 
waiting their chance to swoop down on drying grain. 


[Vol. 42 


ted to cereals and pulses. Greens, such as the tips of peas 
and mustard plants, are chopped, and dried on the cots in the sun, and kept in 
baked clay jars, or in bags improvised from old skirts and shirts. "Mahua" flowers, 
if not all eaten fresh, are similarly dried. Coriander, the most used of spices, is 
gathered by the children and dried by their mothers. Mangoes, peeled and dried, 
are kept in jars to be used later to flavour insipid dishes. Mustard seed is dried 
and stored, ready to be doled out to the oil-presser when a fresh supply of oil is 

The only other forms of preservation within the reach of a village housewife 
are spicing, preserving in oil,, and preserving in salt. Vinegar is not used. The 
containers in each case are clay jars of various sizes, made by the village potter. 
Carrots and radishes will keep about a week, if well mixed with spices and covered 
with oil. The "rai labhera" (small plums) keep longer. And horse-radish, limes 
and mangoes keep for months. The "rai labhera" and limes are scarce, and hence 
available to only a few. They require a little oil, and spice. The horse-radish and 
mango are within every one's reach. Mangoes lend themselves to the greatest 
variety of preserves. In salt they are acceptable, in spices and oil they are a happy 
addition to a noon day meal, and in raw sugar they are a special treat. 

With clay jars or old cloths as her only containers, and a stifling room as her 
only storing space, there it little more that the village woman can aspire to, in 
carrying foods beyond their season. 


Before considering the further processes through which the food materials must 
go in the home, it will be necessary to study the equipment with which a village 
woman works. Her work-room is the courtyard in which all family activities are 
carried on. The earth is her floor, and the sky is her ceiling. From the courtyard 
she sees only these and the walls of mud around her. 

The walls which surround her are made of mud dug from the bottom of a 
near-by pond. She cannot see over the tops of the walls and. more 


he cannot see over the tops of the walls and, more important still, 
over the walls at her. If she is a Brahman, her courtyard will be 
large enough for her to work in, and to spread out her grain comfortably, with 
plenty of sunshine most of the year, and at least a patch of it crossing the floor in 
mid- winter. If she is an outcaste, her courtyard will be small, originally high, but 
her walls will be eroded until low enough to admit abundant sun and air. The 
women who suffer are those whose courtyards are small, and whose husbands insist 
upon keeping them in strict seclusion. Such a one is the oil-presser's wife, whose 
courtyard is cramped, and high-walled. During the rains, she works in puddles, 
and during the winter she shivers between damp walls which shut out the sun. 
She and her children are among the most wretched in the village. 

The carpenters are as near average as any group in the village; and a diagram of 
a carpenter's house is shown on opposite [page]. The enclosure is a typical in that it 
has two rooms off the courtyard. Usually only Brahmans have two, whereas those 


8 Feet = I Inch 


[Vol. 42 


less than Brahmans should be content with one. In this particular case the second 

was added to accommodate extra brothers. It is now falling into disrepair. 



side world. On rare occasions when a village woman is going to her father's home 
for a visit or to the wedding of a relative, she swathes herself in a coarse sheet and 
goes out through the cattle-room. Her only other means of exit is by ladder. If 
her courtyard has a ladder, she climbs to the top of her wall and looks over at her 
neighbour, and if no men are at home in her own or her neighbour's house, she 
drops down into her neighbour's courtyard for a hasty visit. Aside from these rare 
breaks, she spends her life in her own courtyard, with food-preparation as her chief 
task. Her children wear few garments, and these are made by the village seamster, 
and washed by the village washerman. She has no furniture to care for beyond 
the unvarnished roped cots, used as beds by night and as drying-racks by day. 

During the winter weeks, coverings of home spun sheets or padded quilts are 
used. These she throws over wooden pegs projecting from the courtyard wall. 
She has no hangings or rugs, or silver or china to care for. If food processes were 
as simplified for her as they are for us, her life would be easy. But as it is, her 
work is heavy and slow. 

Her activities centre around her "chulha" (fireplace). Generations ago a 
great- grandmother-in-law made it in a secluded corner of the courtyard and each 
daughter-in-law since has kept it in repair. There is a supplementary "chulha" in 
the store-room to be used in stormy weather, but the one in the courtyard is the 
true family hearth. It is made of mud piled up and shaped so as to protect a small 

;ing utensil. (The daily care which the housewife gives 



eating.) Her fuel is the twigs which the children gather from groves, or the dung 
cakes which she and other women of the household have shaped. 

Almost as necessary as her fireplace is her stone flour-mill. All the grains and 
pulses which are to be made into bread must be ground in the home mill. The mill 
is made of two stones, cut like discs about two inches thick and from one and one- 
half to two and one-half feet in diameter. The housewife has embedded the lower 
stone in a base of clay, which lifts it to a convenient height from the floor. Th 
upper stone is free to turn on the lower one, being held in place by a peg in the 
centre of the lower stone which passes through a hole in the centre of the upper 
stone. The clay base has a diameter several inches greater than that of the stones, 
so that it forms a border around the outside of the mill. This is trenched to look 
like a moat surrounding a low stone castle. As the woman turns the upper stone 
round and round with a wooden peg fitted into a slot near the rim, she pours her 
grain into the hole at the centre of the stone. As the grain is crushed and ground 
between the stones it gradually spills out around the edges into the encircling moat. 
When the grinding is finished, the housewife brushes the flour out into a winnow- 
ing basket, and separates chaff and clay-dust from the ground grain. She returns 
some of the bran to the flour, unless the bread is intended for a special feast. The 
grinding is heavy, tedious work, and if possible two women share it, sitting facing 



each other across the mill, each with a hand on the turning peg. One of them 
pours the grain into the centre hole, without breaking the rhythm of the turning. 
The steady hum of the grinding stone is one of the earliest morning sounds of the 
village. In a few homes one finds a "darenti", a mill patterned after the one de- 
scribed, but smaller and lighter. It is used in splitting the pulses. As the supply 
for several months can be food. For the winnowing and sifting there are baskets 
and sieves. Cooking utensils are never borrowed, but the "darenti" is regarded 
differently, although it touches the food. 

Before rice or barley can be ground they must be husked. For this the house- 
wife has a pestle and mortar. The pestle is a thick pole about four feet long, bound 
with iron rings, one of which forms a rim at the bottom. The mortar is a stone 
cup set into the earth floor of the courtyard. The husking is another strenuous 
task usually shared by two women. They sit on the floor one on either side of the 
mortar. Each woman holds a pestle, and turn by turn, each lifts her pestle above 
her head and brings it down heavily into the rice, some of which is struck, but 
much of which flies out of the mortar. One of the women keeps brushing the 
scattered grains back into the mortar, while managing to keep her hand out of 
the way of the pestles, and not missing her turn. The grain thus treated is not 
only husked but chipped and broken. However, skilful winnowing saves every 
scrap that might [be] food. For the winnowing and sifting there are baskets and 
sieves of several sizes and shapes. The winnowing baskets are of reeds. The sieves 
may be strung with coarse threads or reeds. 

The utensils which the village housewife uses are made of brass, iron, stone, 
clay or wood. When she makes her daily bread (she actually does make it every 
day) , she may mix and knead her dough in a bowl of brass or wood, or baked clay. 
She pats out the flat round cakes with her hands if the flour comes from mixtures 
of corn, barley, millets or pulses. If the cakes are of wheat, and especially if they 
are intended for guests, she rolls them out thin on her small round breadboard, 
which is of stone rather than wood if she can afford it. To bake her bread, she 
may rest over her fire a convex iron plate which she brushes with clay, to keep the 
cakes from sticking. Or she may fry the cakes in deep fat in an iron vessel shaped 
very much like those which we use for deep-fat frying. She lifts the cakes out of 
the fat with a long-handled, flat-bowled iron spoon, which may or may not be 
perforated. And when her bread is ready and her husband calls for it she serves it 
on a brass tray, always brightly polished. For cleaning her brass she rubs loose a 
bit of her earth floor, applies it with a knot of grass and much rubbing, and rinses 
it off. 

When she prepares vegetables she rarely removes the skins, but cuts up the 
whole vegetable with a small sickle. She braces the handle under her toes so that 
the blade stands upright. Against this she pushes the vegetable. Greens are held 
<in bundles and cut through in the same way. To fry her vegetable, she uses the 
same iron pan for deep-frying. To boil her "greens" (leaves and herbs), she uses 
a brass kettle with straight sides and no handles. Both "greens" and vegetables 


[ >L. 42 


she serves on the brass tray with the bread. The bread is broken off bit by bit and 
used to dip up the other food. If she is serving guests, she will place bread and 
vegetables and any other food on clay saucers or dishes made of large leaves bound 
together with twigs. These can be thrown away when the guests depart. In no 
case should household utensils or dishes be touched by others than members of the 
family. When I admire a piece of brass, I always admire from afar. 

Every day the housewife grinds the spices to be used in the food. For this she 



gracefully to a point at one end. On its corrugated surface she lays her 
spices, sprinkles them with water, then rolls a small, rough stone roller back and 
forth over them until they form a paste. 

Every morning and evening she puts fresh milk on to simmer. For this she 
uses a baked clay jar, set in a smouldering fire of dung cakes. A separate fire- 
arrangement is required for this long period cooking. It consists of a large, shallow 
basin with a thick base which she has made of clay. In it she places a burning 
dung cake, with others arranged over and around it, like a nest for the jar of milk. 
After hours of simmering she pours the scalded milk from this clay jar into another 
kept for the purpose. And the next morning she pours the clabbered milk into 
still another clay jar which serves as a churn. Each of these jars has a distinct 
'shape and is called by a different name. The churning paddle is a stick of wood 
with cross pieces at the bottom. The woman who is to churn loops a rope around 
the leg of a cot and then twists it at the middle around the paddle, and holds the 
two ends in her two hands. By pulling one end of the rope and then the other, she 
rotates the paddle horizontally and keeps it rotating until butter forms. The 
churned butter is placed in a special jar with other butter and accumulated for a 


The utensils mentioned, along with the large clay jars which are filled each day 
with water, our housewife considers necessary to the feeding of her family. With 
increasing means and increasing family, she may add to the number of these same 
'articles, or she may replace clay with iron, or iron and wood with brass. Or she 

ng utensils and serving dishes of varied sizes and shaoes. She 

may add brass cooking utensils and serving dishes of varied sizes and shapes, 
may aspire to a serving dish or cup of bell metal, which is more expensive than 
brass. Beyond this, her desires do not go. 

The following list of utensils indicates the variety used, and the relative im- 
portance of the different articles, as shown by their frequency. The list includes 
forty-eight households which represent all castes and all degrees of prosperity. I 
have divided them roughly into what our village friends call "rich", "moderate", 
and "poor". ^ They would add to the list "very poor", but these I have included 
under "poor". There is no way of estimating the actual wealth of a family. Those 
who appear poor are very apt to have money and jewellery hidden away in a wall 
or the ground. Indebtedness is no register, as most of the village is in debt: 




»/ articles, with percentage of rich, moderate, and p 

in which each occurs 

Classification of homes 

Local name 


5 rich 

22 mode- 

f% < 


21 poor 

Per cent. 

Per cent. 

Per cent. 

Chakki . . 

Grinding stone 




Darenti . . 

Light mill for splitting pulses . . 




Musal . . 

Pole-pestle for husking rice . . 





Stone mortar for rice . . 




Charpai . . 

Roped cot on which foods are 




Chalni . . 

Round wooden sieves strung with 




Sup . . 

Reed winnowing basket 




Parat . . 

Brass mixing bowl 




Tasla . . 

Iron mixing bowl 



• • 

Kanthani . . 

Wood mixing bowl 




Katheli . . 

Smaller wood mixing bowl 




Kunra . . 

Clay mixing bowl . . 



m • 

Kunri . . 

Smaller mixing bowl 




Pata (w) . . 

Wooden rolling board 




Pata (s) 

Stone rolling board 





Wooden rolling pin 




Tawa . . 

Convex iron griddle 




Karahiya . . 

Iron pan for deep fat frying . 




Hansiya . . 

Small sickle for cutting vegetables 




Sil batta . . 

Stone on which spices are ground 




Lorhwa . . 

Roller for grinding spices 




Batua . . 

Brass kettle for boiling greens . 





Small cooking pot 




Mathni . . 

Churn . . 




Chamcha . . 

Long handled iron spoon 




Kalcha . . 

Different type of iron spoon 




Chamcha (wood) 

Wooden spoon . • 

• • 

• • 


Chimta . ♦ 

Iron fire tongs • . 




Chimti . . 

Smaller iron tongs . . 



• • 

Sansi . , 

Tongs curved for lifting kettles . 



• • 

Thara . . 

Clay water jar . . 




Kalsa . . 

Brass water jar • • 



• « 

Chilamchi . . 

Brass wash basin . . 


• • 

• • 

Silafchi . . 

Brass basin • • 


• • 

• - 


Brass serving tray . . 




Phul Thali 

Tray of bell metal . . 




Rakabi # ♦ 

Brass plate . . 



• • 

Lota , , 

Brass drinking bowl 




Phul lota 

Drinking bowl of bell metal . 


• • 

• » 


Brass cup in which food may be 




Phul bella 

Bell metal cup 




Gilas . . 

Brass glass . • ♦ • 




Ghanti . . 

Small brass water vessel • • 




Tarazu . . 

Balance scales 




Pandan . , 

Box containing betel condiments 




Clay saucers . . 

Cups and jars . • - • 100 




[Vol. 42 



Every day the village woman makes fresh bread for her family. As has been 
mentioned, she makes it of wheat, barley, or mixtures of these with pulses, from 
the spring harvest until the end of the rainy season. From the end of the rains 
until spring, she makes it of maize, different millets, and perhaps rice, with a wheat 
mixture occasionally as a treat. 

Whatever the grain, she suns it well the day before she uses it. Early on the 
day when she is to bake it into bread, she, with the help of another woman of the 
family, grinds and winnows it. She mixes the flour with water in her mixing bowl 
and kneads it by alternately working it vigorously with both hands, and pounding 
it with her knuckles. When she considers it "soft enough," she takes up a ball of 


the dough and works it further in her hands, before patting and rotating it into a 
round, flat cake. While she turns the cake in her hands, she pinches the edges. 
She brushes a daub of wet clay over her thin, convex griddle, to keep the cakes 
from sticking. As each cake is shaped, she flops it on to the griddle. There is only 

for one cake at a time. She turns it over with her long, flat iron spoon, 
the first cake cooks, she shapes the second. As soon as the first is done well 
on one side, and slightly on the other, she stands it upright inside the fireplace 
under the griddle, to toast, while the second cake cooks above, and she shapes a 
third. Meanwhile, she must keep her fire alive by blowing it and adding fresh 
fuel. Her fuel may be small twigs, gathered by the children, or dung which she 
has gathered in her stable, or which the children have collected in fields or on 
roads, and which she has mixed with chaff and water and shaped into cakes of con- 
venient size for her fireplace. The bread which toasts close besides the fire, must 
be watched lest the burning fuel touch it, or lest it topple over into the flame. If 
its edges have been pressed together properly, the cake swells out like a balloon, as 
the moisture inside turns to steam. At this point, the cook lifts it out with her 
iron fire tongs, and bangs it against anything convenient to drive out the expanded 
air. The steps in bread making are listed on page 335. 

The women prefer wheat, or mixtures in which wheat appears, not only because 



rolling-pin, after patting them between her hands. Cakes of other grains fall 
apart if she tries to make them equally thin. Corn cakes are a cross between corn- 
meal mush, and corn bread. Cakes of the little millet look like clay, before they 
are baked. They are mealy and have a raw taste. Cakes of sorgum are brown 
with a purplish tint, and are slightly bi 
Poorman's millet are still less desirable, 
to be fried in deep fat. When fried, the 

Cakes of Italian and of 


fried, the bread is known as "puri" as distinguished 
from the toasted bread which is "chapati". Both "puris" and "chapatis" are 



covered by the general term "roti". The occasions on which "puris" are required 
are discussed under the ceremonial significance of food. 

Parched grains provide a useful variation in village fare. The grain-parcher is 
•kept busy most of the day when a fresh crop of grain is cut. At other times he 
keeps his furnace going late each afternoon. The farmer or one of his children 
brings the grain to be parched. The parcher has a clay furnace in which he burns 
leaves and twigs. In the furnace he heats clay jars of sand. He mixes the grain 
with the hot sand and keeps mixing them until the grain has popped or puffed, or 
browned. Then he sifts out the sand, and gives the grain back to the farmer, after 
taking out a portion as his pay. The grain thus roasted may be eaten as it is. Or 
the farmer's wife may grind it coarsely for his morning meal. If there is butter- 
milk, it is added to the ground, parched grain. If not, water is added, and a little 
raw sugar, if there is any. If not, there is always salt. The daily ceremonies which 
must precede any cooking make it impossible to prepare an early cooked breakfast. 
This ground, parched grain, called "satua," solves the breakfast problem for the 
farmer, and perhaps for the children. The women and children usually eat cold 
food left from the day before, but the men scorn this. 

If the grain to be parched or roasted for "satua" is fresh, it goes straight to 
jthe parcher. But if it is dry — as it is most of the year — the farmer's wife must 
soak it for twelve hours, and sun it for six hours before her husband carries it to 
the parcher. Corn, barley, or a combination of barley and gram or of barley and 
peas, are the most popular. A quantity is sent to the parcher that will supply the 
family with "satua" for several weeks, or perhaps, months. A prosperous farmer 
may carry a maund and a half (about one hundred and twenty pounds) to the 
parcher's, while his poorer neighbour carries about ten pounds. And each hopes 
to have "satua" enough to feed himself for the same period. When the parched 
grain is brought home, the women pound it in the stone mortar set in the court- 
jyard floor. This loosens the outer skins. Then they clean it, and grind a week's 
supply at a time. 

At harvest time there is a feast of parched foods in which the parcher has no 
share. While men rest from harvesting in the fields, or while they sit with their 
brotherhood after a strenuous day's work, they gather a pile of leaves or stalks and 
scatter the heads of green grain or the pods of legumes within the pile. They light 
it from a scrap of smouldering cow-dung, always on hand for lighting pipes. When 
the blaze dies down, they scatter the charred remains and pick out the smoking 
heads of grain. These they rub between their palms until the kernels are freed. 
Some are still green; others are charred black. But the combination is very agree- 
able. The women and children manage to get a share of the heads for roasting, 
from the stalks which the men bring home at the close of each day's work in the 
harvest fields. Peas, wheat, and barley are thus treated and eaten in March and 
early April. During the harvest at the end of the rains, the heads of both large 
and small millets are laid on the fire, and the green ears of corn are husked and 
turned in a blaze or in the embers of a fire of twigs. The corn is field and not 

[Vol. 42 


sweet corn, but it is good roasted. These popular products are available to everyone 
in the village, for at these seasons every able man has a chance to help some farmer. 
And those who cannot, plan to be present when generous neighbours indulge. The 
amount of parching done in the fields depends on the good will of the owner of 
the field. If he is miserly, or work is pressing, he may allow little time for rest 
and feasting. But ordinarily, it is an accepted part of the day's routine. 

Other dishes in which cereals appear, are prepared for certain festivals. These 
come so seldom that the foods are not of great significance from the point of view 
of nutrition. In the eyes of the village housewife they are tremendously important 
and consume much of her conversation and time. They are at the end of this 
chapter, along with other festive dishes. 

Rice is treated differently from the usual cereals. For bread it must be husked 
with mortar and pestle, then ground. It makes a very soft dough. The cook 
keeps her hands moist while handling it. It is toasted like any other bread. Rice 
is used for bread only when there are no other grains on hand. 

It is more often used in "khichri." Water is boiled in a brass kettle, and rice 
and a split pulse are dropped in together and cooked. They may be in equal pro- 
portions, or there may be less pulse. 

A variation of this with a decidedly different flavour and "feel" is "dal bhat." 
The rice and split pulse are cooked separately and combined when almost dry. The 
women of the family drink the rice-water. 

Sometimes rice is soaked and later fried in "ghi" (clarified butter) and salted. 
Oil may not be used for this. "Khir" is a dish which is limited to households 
where there is enough milk so that all is not needed for "ghi." Rice is boiled in 
milk and sweetened (one-half pound rice to about three quarts milk). Villagers 
say that there should be as much raw sugar added as there is rice. Just before 
serving, a little "ghi" should be added. 

A variation of this dish, which is available to a much larger number, is 
"Raskhir." The juice of the sugarcane is boiled, and rice cooked in it. It has a 
flavour which we think disagreeably strong, but village folk eat quantities of it 
with pleasure. A cross between these two, is a dish called "maheri," in which 
buttermilk and raw su^ar are *AApA try rM\r*A *-tV* 


But in 

appears at the time 

sand. Every 

better variety of rice necessary 

Other occasional uses 










Weight before 

Weight after 

Bringing dung from drying pile and bring- 

20 minutes. 

ing water. 

Pounding with pestle 

3 minutes . . 

3 lb. . . 3 lb. 

Winnowing, to remove insects, husks, etc. 

2 minutes . . 

3 lb. 

2 lb. 13 oz. 


1 hr. 5 min. 

2 lb. 13 oz. 

2 lb. 9 oz. 


2 minutes . . 

2 lb. 9 oz. 

2 lb. 8 oz. 

Making into cakes 

45 minutes 

Water+2 lb. 
8 oz. 

4 lb. 3 oz. 


Pounding with pestle and winnowing to 

30 minutes. 

3 lb. 

2 lb. 12 oz. 

remove insects, husks, etc. 


1 hour . . 

2 lb. 12 oz. 

2 lb. 10 oz. 


3 minutes. 

2 lb. 10 oz. 

2 lb. 9 oz. 

Kneading, rolling, cooking . . 

• • 

Water+2 lb. 
9 oz. 

4 lb. 

Number of cakes mac 

ie from one pound of grain 



• • 



• • 


Large millet (sorgum) . . 

• • 


Small millet . . 

« • 


Tirra (wheat, barley, gram) 

1 • • 


Gojai (wheat and barley) 

• • 


Rice • • • • 

* • 



• • 


Bejai (barley and peas) 

• • 


Wheat — grain . . 

• 9 


WhMt (c-xVes. $r\eA in deeo 

fat} . . 


%c. clarified butter. %c. left after frying. 




our fC dal roti" (split pulse and bread), which is his idea of a satisfactory meal. 
The term "dal" is applied primarily to the dish of cooked legumes served with 
bread. It also applies to any of the pulses, after they have passed through the 
splitting process and are like split peas. Before this stage there is no inclusive class 
name, and each pulse is known by its plant title. If turned to some purpose before 
being split, a special descriptive term is used for the particular product. For 
instance, split peas may be referred to simply as "dal". If cooked green, they are 

* * m * ft 



desired "dal" comes, and his wife in taking "dal" out of storage for some particular 
dish, is careful to select the split product appropriate for her purpose. For the 

[Vol. 42 


ordinary "dal" to be cooked and eaten with bread, she has no choice beyond the 
supply on hand or her husband's digestive powers. The uses of the pulses, before 
and after they are split, are outlined at the end of this section. 

When "dal" is to be served, the housewife gets out the desired amount the day 
before. If she has "mung", "urd" or "moth" she pounds it with her pestle, then 
soaks it overnight. By morning she can rub off the loosened husks under water 
and skim them from the top. She does not throw the husks away, but saves them 
and combines them with barley and wheat in making bread. If she is in a hurry, 
she may omit the soaking and let the family eat husks along with the pulse. Field 
peas do not require soaking, as husks come off during the splitting. She heats 
water (about two quarts to a pound of "dal") and while it is heating, she prepares 
her spices. She grinds red pepper and half a root of turmeric on her spice-stone, 
and adds them to the boiling water. Then she adds the "dal" and salt (one tea- 

ooks it until it is pulpv when rubbed 

thirty minutes). Whe 

(about twenty 

seeds) and adds it to the "dal". If she has cinnamon, cloves and cardamom, she 

adds them as well. 


clarified butter on the tip of the big iron spoon, and drop a few chips of asafoetida 
into it. She heats this until it sizzles, then stirs it into the "dal" and covers it 
quickly with a wooden lid. Our village friends assure us that the asafoetida 
prevents flatulence. 

The "dal" is quite soupy, but any one brought up in the village knows how to 
eat it efficiently, if not noiselessly, by dipping it up with pieces of bread. We like 
the flavour of all the "dais," but "mung", "urd", and pigeon pea have a slippery 
feeling which is unpleasant. Cooking the husks with the pulse, relieves this some- 
what. Field peas may be cooked as these other "dais", making a more appetizing 
dish, at least for us. 

Certain "dal" dishes remind one of meat substitutes, although to the villagers, 
meat is our "dal" substitute. One such dish is "mungauri" in "jhori". To make 

\e farmer's wife takes one pound of "mung" from her store, and soaks it to 
the skins. She rubs it between her hands and washes it until the pulse is 
clean. She then grinds it on her spice-stone, handful by handful, until it is a paste. 
To this mushy product she adds a teaspoon of salt. Into her pan for deep-frying 


l A 

If it is mustard oil, 

she heats it until it becomes clear and the odour is gone. Then she drops the 
"mung" into the hot oil or "ghi". She does this by taking up a handful and shoving 
off a bit at a time with her thumb. It forms irregular lumps about three-fourth 
inches in diameter, which she removes from the hot fat when brown. These may 
be eaten as they are, but should be served in "jhori". She makes the "jhori" by 
adding water to more "dal" paste and frying the mixture in a tablespoonful of 
"ghi". For flavour she adds a few seeds of "methi" (fenugreek). She stirs it 
occasionally, but does not scrape loose that which sticks to the pan. She then adds 
turmeric and salt to taste, and still later adds coriander, black and red pepper, one 



cardamom, three cloves and a small stick of cinnamon, all ground together. At 
first this "jhori" is an unappetizing, brown soup somewhat lumpy. The spices 
turn it to a lighter brown and further cooking makes it a thick gravy. When it 
is done, the fried cakes of "mungauri" are dropped into it. The product is very 
much like hamburg cakes in gravy. These same cakes, "mungauri", can be cooked 
with spiced potatoes, in which case it seems still more like a meat dish. 

"Mithori ka jhor" is a similar dish. When a woman makes it, she cleans her 
"mung", then grinds it like flour with her heavy grinding stone. She adds a touch 
of asafoetida and red pepper, and makes it into a thin batter. This she drops bit 
by bit into a kettle of boiling water. As soon as the little cakes or lumps are firm, 
she takes them out and dries them on a roped cot in the sun. She can put them 
away and use them from time to time to vary her pulse dishes. When she wishes 
to serve them she browns them in oil or "ghi", and adds buttermilk, thickened with 
a flour of gram — one-fourth pound of flour to two- thirds quart of buttermilk. 

Still another dish on the meat-cake order is made by preparing and frying small 
flat cakes of gram flour, and adding buttermilk. The cakes are "danre" and the 
gravy "karhi". It tastes rather sour, an unfamiliar flavour to us in gravy. It 
should be eaten with rice or bread of some wheat mixture. The "danre" of gram, 
is crumbly, whereas the "mungauri" made of "mung" is slightly tough. 

Pulses may be combined with cereals in bread. They make a stickier dough 
which is harder to handle than a dough of wheat or corn. They also add a distinct 
flavour of partially cooked peas. A favourite village dish is a wheat cake with 
pulse filling. To make these cakes the housewife makes a dough of wheat flour as 
for bread, but adds more water to make it softer. To one pound of wheat flour, 
she makes a filling of six ounces of "urd", soaked and cleaned and rubbed on her 
spice stone to a paste. She spices it with red pepper, three cloves, a stick of cin- 
namon, forty coriander seeds, one cardamom and twenty seeds of aniseed, a touch 
of asafoetida, and a teaspoon of salt. She takes a ball of dough in her hand, flattens 
it out, lays some of the filling on it and draws up the corners to cover the filling. 
Then she rolls it out, like a wheat cake, and drops it into hot fat. It puffs and 
browns, but loses none of its filling if properly prepared. When finished it makes 
seventeen cakes, weighing two pounds. When eaten hot, these filled cakes are 
delicious. The difficulty with most of these pulse dishes is that they should be 
prepared immediately before serving. This is one reason why village feasts are 
always several hours late. 

Although pulses are used chiefly as "dal", dried and split, they make a distinct 
contribution before reaching this stage. The tender tips of the growing plants of 
field peas are cooked as greens, or they may be cooked with potatoes as a regular 
vegetable dish. Small potatoes are cut in half and browned in "ghi", which is 
flavoured with asafoetida. A little water is added, along with salt, red and black 
pepper, cloves, coriander, and cardamom. The pea tops are added and the whole 
is cooked until the potatoes are tender. Green pea pods may be cooked with 
potatoes in the same way. 

[Vol. 42 


The long pods of the cow pea may serve as a vegetable. If they are green 
enough, they are cut up with the sickle, browned in a little oil or "ghi" with 
turmeric, black and red pepper, cloves, cinnamon, cardamom, coriander and salt. 
Then a little water is added, in which the cut up pods are cooked until tender. At 
the last a thin paste of gram flour and buttermilk is added. If the pods have begun 
to dry, they must be removed and the peas cooked without them. 

The cluster bean is chopped up and cooked in the same way, but without any 
addition of buttermilk. Most of the water is boiled away. Both these cluster 
bean and cow-pea can be dried in their pods and stored, to be used later as a vege- 
table dish. 

Like the cereals, pulses are popular when roasted, or parched. Sometimes they 
are roasted in their pods while green, but ordinarily they are first shelled and 
dried. When parched they may be eaten just as they are, like the gram which 
carries many a traveller through days away from home. Or they may be ground 
coarsely with parched cereals into "satua" for the farmer's breakfast. 



as we eat fruit. During the carrot season one sees the children, especially in the 
quarters where vegetable growers live, gnawing at raw carrots while they creep 
after their mothers or play together in the lanes. The same is true when sweet 
potatoes are plentiful. The sweet potatoes may be partially roasted in a bonfire 
built in the lane or on the courtyard floor. And in the hot weather, men, women 
and children may be seen feasting on long green cucumbers. Roadside vendor stalls 
and religious fairs abound in cheap cucumbers so that one cannot miss them, at 
home or abroad. 

The vegetables which grow less abundantly are cooked and eaten with bread. 
When there is a vegetable dish, the pulse is usually omitted, unless there is great 
prosperity in the home, or a guest. Practically the same plan is followed in pre- 
paring all vegetables. Seeds and other inedible portions are removed. Most skins 
have been found edible, and remain. Potatoes are never peeled. Neither are 
plantains. The vegetable is cut up with the small sickle mentioned. Every village 
girl learns to brace the handle of the sickle firmly with her toes, and press the vege- 
table against the blade, one hand on either side of the blade. 

A tablespoonful or two of "ghi" where there are milch animals, or mustard oil 
where there are none, is heated with a little fenugreek or in some cases asafoetida, 
in the pan for deep frying. When the fenugreek is well browned, the chopped 
vegetable is added, with enough water to keep it from burning. While it cooks 
spices are added, varying from salt, pepper and coriander in poorer homes, to these 
plus cloves, cinnamon and cardamom in prosperous families. The yellow-green, 
spice-coated, oily product is partly mush and partly in lumps. None of the liquid 
is poured off. It is eaten with the help of torn-off bits of bread. In appearance 
it may not be appetizing, but it is delicious. 



amounts used 

Pigeon pea 

More than any other 
as "dal". But less 
than gram for total 




used is more 
other pulses. 

Field pea 

Less than gram or 
chick pea. More 
than urd, mung, 

or moth. 

Pulses used in the homes of Karimpur 

Uses when not split 

1. Roasted 




Green pods, as vege- 

Roasted plain. . . 

Soaked, fried in "ghi", 
salt and pepper. 

Roasted. Ground with 
wheat and barley, or 
wheat and peas for 



1. Tops of plants as 

2. Green pods and green 
peas as vegetables. 

3. Dried, used as vege- 

4. Roasted in pods. 

5. Roasted. Ground with 
barley, or barley and 

gram for 


Preparation of split product — "Dal" 

Processes at 
time of harvest 

Processes at 
time consumed 

Soaked, dried. Split. 
Winnowed. Fried fur- 
ther. Stored. 

Half split. Winnowed. 
Dried further. Stored. 
Other half similarly 
treated in September 
for remainder of year. 

Stored unsplit. Later 
enough for 3 or 4 
months split at a time. 

Pounded (mortar and 
pestle). Winnowed. 

Pounded. Winnowed. 


Use of split product 


1. Boiled with spices. 

2. Ground. Made into bread, 
alone, or with wheat and 
barley, wheat and gram, 
or gram alone. 

1. Boiled with spices. 

2. Ground into flour, to be 
used in special sweets. 

3. Ground and made into 
bread, alone, with pigeon 
pea, with wheat and bar- 
ley, or barley. 

1. Boiled. Only eaten in this 
way when other "dais" 
are gone. 

2. Ground, with barley, or 
wheat and barley, to make 



Pulses used in the homes of Karimpur — (continued) 


amounts used 


# • 

Less grown than peas. 
More used as "dal." 

Uses when not split 


Slightly less than urd. 

Roasted and ground for 



« « 

1. Roasted. Ground for 

2. Roasted alone or with 
maize and ground for 

morning "satua." 

Preparation of split product — "Dal" 

Processes at 
time of harvest 

Rubbed with oil. 

Rubbed with oil. 



Rubbed with oil. Split 

Processes at 
time consumed 

Pounded. Soaked. 
Husks rubbed off. 

Pounded. Soaked. 
Husks removed. 

Pounded. Soaked. 
Husks rubbed off. 

Uses of split product 






Boiled with spices. 
Cooked with rice. 
Cooked separately and 
added to rice. 
Ground with wheat or 
wheat and barley for 


Ground for special cakes. 





Boiled with spices. 
Like urd, above. 


Combined with wheat, 
or barley, or both, and 
ground for bread. 
Ground for sweet-cakes. 

Ground and used in meat- 
like dishes. 




Cooked with spices. 
Especially for invalids. 
Mixed with other pulses 
and millet, and ground 
for bread. 

Combined with "mung", 
in meat-like dishes. 



| Little used by men, 
More for animals, 


String beans • 


Small amounts grown 
by Kachhi vege- 
table growers. 

1. Cut up fresh. Cooked 
as vegetable. 

2. Dried. Stored. Soaked. 
Cooked as vegetable. 

1. Cooked fresh. 

2. Dried. Stored. Soaked. 
Cooked as vegetable. 

1. Cut up. Strings re- 
moved. Cooked as 
vegetable, usually with 



[Vol. 42 


Potatoes are a popular base for this vegetable dish during the autumn and 
winter months. And almost any vegetable may be added to them. Egyptian arum, 
similar to dasheen corms, is available as potatoes decline, and it makes a satisfactory 
base. The vegetables used in this conventional vegetable dish, either with potato 
or Egyptian arum, or with each other, or alone, are carrots, radishes (alone or with 
tops), sweet potatoes, cauliflower, cabbage, egg plant, okra, yam, tops of young 
pea plants, field peas and cow peas with or without pods, string beans, and cluster 
beans, and all the varieties of cucumbers and gourds available. Mushrooms are 
sometimes included. And fruits, such as mangoes, plantains and melons may be 

used in this way. 

"Reota" is a dish in which the village cook combines vegetables and buttermilk. 
She may make it of white goose-foot, potato, pumpkin or gourd. She boils the 
potherb or vegetable, then cooks and drains it. She wraps a bit of asafoetida in 
cotton, and sprinkles oil on it. This she lays in a smouldering dung cake, and over 
it she inverts the vessel in which the "reota" is to be prepared. This vessel should 
be of baked clay, to absorb the fumes better. While the asafoetida fumes, she rubs 
the vegetable on the spice-stone to a paste. Next she uses the stone for the necessary 
spices — small red pepper, one seed of black pepper, one teaspoonful coriander seeds, 
one clove, and about three-fourths teaspoonful salt. Then she mixes the spices in 
with the vegetable, and rubs them all together on the spice-stone. She sets the clay 
jar upright and pours buttermilk into it, and puts on the wooden lid quickly. Then 
she scrapes the vegetable and spices into it, and again puts on the lid. She serves 
it cold. It looks like a mash, and is quite attractive when green vegetable is used. 
When one becomes accustomed to the flavour, it is very welcome in hot weather. 

Similar to "reota" is the vegetable dish in which buttermilk and gram flour 
are blended and added to green peas or beans which have been cooked until tender. 
One other successful vegetable dish is made from the leaf of the Egyptian arum. 
It is not prepared as much by village women as by those in town. The large leaf 
is opened out and covered with gram flour and spices. Then it is rolled up like a 
jelly-roll and sliced off as a roll is sliced. The round, filled slices are fried in "ghi" 
or oil. It is one of the best tasting of vegetables. 

The following table (pages 342 and 343) gives the ordinary procedure followed 
in preparing vegetables. Vegetables representing various classes and various 
methods of treatment have been chosen. To us, the differences in method are 
slight, but to the village cook they are very important. 

"Achar," which approaches our pickles, is made from carrots, radishes, beans, 
and horse-radish, and onions, in low caste homes. When a farmer brings in a 
generous supply of any of those, his wife uses a portion for "achar." She cuts up 
the vegetable as for the ordinary vegetable dish. She scrapes and washes carrots 
and onions. She strings the beans and removes the tough fibrous skin of the horse- 
radish. The last two she cooks. When prepared, cooked or uncooked, she spreads 
them in the sun to dry. While they dry, she grinds on her rough spices tone: husked 
mustard seed, coriander, black pepper, red pepper, turmeric, with perhaps cloves 



and big cardamom. She grinds these dry, not moistened as she does for the usual 
vegetable dish. She puts both vegetables and spices in a clay jar and stirs them 
together well. Last, she adds enough mustard oil to cover the vegetable with a 
spicy, oily coat. Beans are ready in twenty-four hours. The others she must shake 
and sun daily for four or five days. They must all be eaten soon after they are 
ready, as they rapidly ferment. 


Certain vegetables are boiled in water in a deep kettle, without the addition of 
fat. When prepared in this way they are designated as "sag" — greens. Potherbs, 
tops of growing pulse or mustard plants, radish tops, leaves of Egyptian arum, 

dill, and "palak," a leaf resembling our spinach, may all be cooked as "sag." 
During a few months, fresh green leaves are available. Some of them are dried 
and stored for later use. 

When a village woman plans to cook "sag," she cuts up the leaves by pushing 
a bundle of them against the upright blade of her sickle, layer by layer. She has 
learned to cut them very fine in this way, without slicing her fingers with them. 
She then washes the leaves and cooks them in water in a straight, deep brass kettle. 
She uses one quart of water to one pound of vegetables. The water is warmed 
and salted before she adds the vegetables. After it has begun cooking, she adds 
"dal" or "mung" or "urd" (one-third cup to one pound vegetables). She is care- 
ful to add it so that both "dal" and leaves will be finished at the same time. To lend 
flavour she adds several slices of dried mango. When she is ready to remove the 
kettle from the fire, she adds a small amount of wheat flour (about one-fourth 
cup) slowly while stirring. She explains that "without flour the 'sag' will be by 
itself and the water by itself." The flour binds the whole. As soon as she removes 
it from the fire she heats a teaspoonful of "ghi" and a broken up clove or a chop 
of asafoetida, in her iron spoon over the flame. When it is sizzling she plunges it 
into the "sag" and covers it quickly with a wooden lid. 

An old village woman once taught me this song: 

"O daughter-in-law, today I want the leaves, the tender leaves plucked from the tops of 

young grain in the field. 
Tomorrow it will be fenugreek that I require. 
On Wednesday I myself shall nip the pale mustard tops and you will cook them delicately 

in oil. 
On the day following a wild potherb will do. 
On Friday still another, much more satisfying to the hungry — the yam — Don't forget. 

And after that, egg plant will follow, stewed and spiced. 

The squash or perhaps the bitter melon, we shall leave for Sunday. 

Always remember, O wife of my son, I have no taste for bitter melons, often." 


Most of the wild fruits, and berries of the peepul and neem trees are eaten raw, 
and usually when hard and green. Very few of them reach the women in their 
courtyards. Melons are eaten raw when ripe. When green they are cooked as 
vegetables. The local plantains are a small, inferior variety, and are preferred 

Methods of Preparing Vegetable Dishes 

Preliminary process 

Potatoes: Cut up. Skins not 
removed. Washed after 
being cut up. 

Cooking procedure ordinarily followed 

Fat in kettle, with fenugreek, heated. Potatoes 
added, and eriough water to show at edges. 
Other spices. Cooked until potato tender. 
Any water left is served with it. 

Egyptian arum: Boiled. Skins 
removed. Water thrown 
away. Chopped fine. 

Egg plant: Cut up. Washed. 

Mustard tops: Chopped up. 
Washed. If dried — soaked. 

Radishes: Cut up, leaves and 
radishes. Washed. 

Fat heated in kettle with fenugreek. Arum 
and little water added. Spices added. 
Cooked until tender, and water boils away. 

Fat and fenugreek heated in kettle. Egg plant 
added and a little water. Cooked until 
tender. Water boils away. 

Fat and asafoetida heated in kettle. Mustard 
and very little water added. Other spices. 
Water not poured off. Served with it. 

Radish seed pods, unripe. 
Cut up into small pieces. 

Fat and asafoetida heated in kettle. Radishes 
and tops added. Small quantity water. 
Spices. Cooked until radishes tender and 
water boiled off. 

Water Chestnuts: 


ins re- 

Fat and fenugreek heated in kettle, 
added. Also cut up potatoes, and 
enough to show. Spices. Cooked 
potatoes tender, and no water left. 



Fat and fenugreek heated 
added with little water, 
until tender. Water left 

in kettle. Nuts 
Spices. Cooked 
is poured off. 

Fenugreek. Black and red 
pepper. Coriander. 

Fenugreek. Pepper, 
ander. Turmeric. 


Fenugreek. Dried Mango, 
Coriander. Pepper. 


Pepper. Cori- 


Pepper. Cori- 

Fenugreek. Pepper. Cori- 
ander. Low castes — onion. 

Fenugreek. Cardamom. Pep 
per. Turmeric. Coriander, 

Variations in method 

Boiled, peeled, mashed. Cooked in 
fat in kettle with usual spices 
plus turmeric. Buttermilk added. 
Cooked fifteen minutes. 

Same as ordinary, except 
termilk is added in 

that but- 
place of 


1. Like ordinary, combined with 

2. Like ordinary, with gram added. 

1. Like ordinary, combined with 

2. Boiled as "greens." 

1. Pickled. 

2. Tops cooked as "greens." 


Turayan: Skins removed. Cut 
up. (Kind of gourd). 

Lauka (kind of pumpkin) : 
skin scraped off. Cut up. 
Seeds removed. When green, 
skin and seeds remain. 

Cluster bean: Cut up. Washed 
(green in pods). 

Cauliflower: Cut up, flowers, 
and stalks all together. 

Cabbage: Cut up. Washed. 

Karela: (Bitter melon). Skin 
scraped off. Washed. Seeds 

Aniseed and fat heated in kettle. Gourd 
added. Little water. Spices. Cooked until 
tender. Water left is poured off. 

Aniseed and fat heated in kettle, 
and little water added, 
added. Water boils away. 

Cooked. Spices 

Asafoetida and 
added and a 
until tender, 

fat heated in kettle. Beans 
little water. Spices. Cooked 
Water remaining served with 

Fat and fenugreek in deep brass kettle — not 
frying kettle. Cauliflower added. A little 
water. Spices added. Cooked until tender. 
Water kept. 

Fat and fenugreek in kettle. Cabbage added. 
Little water. Spices added. Cooked until 
tender and water boiled away. 

Melon stuffed with spices, plus dried mango, 
and onions (if low caste). Partially fried, 
then water added. Cooked until tender. 
Water remaining, poured off. 

Turmeric. Coriander. Ani- 
seed. Pepper. 

Aniseed. Pepper. Coriander. 


Pepper. Cori- 

Fenugreek. Blackpepper. 
Cloves. Coriander. Tur- 
meric. Cardamom. Cin- 

Fenugreek. Blackpepper. 
Cloves. Coriander. Tur- 
meric. Cardamom. Cin- 
namon. Low cast 

Coriander. Pepper. Fennel. 
Mustard. Asafoetida. 

1. Same as ordinary, combined with 
Egyptian arum. 

2. Or cooked and mashed in but- 

1. May add gram. In this case, is 
only eaten by immediate family. 

2. May be added to buttermilk. 

Same as ordinary, combined with 

When cooked with potatoes, usual 
frying kettle is used. Red pep- 
per and turmeric mixed with 
vegetables before cooking. Other 
spices at end. 

Same as ordinary, combined with 

May add gram to filling 

[Vol. 42 


cooked, in combination with vegetables. "Mahua" blossoms are eaten raw by the 
children. In case they are unusually plentiful, some are dried and cooked with 
vegetables later. "Bel" fruit and the wood-apple are usually roasted over a home 
fire until they burst. Pomegranates are rare. When they do appear, they are eaten 
raw. Guavas are eaten raw, usually when green. Wood-apples are also eaten raw. 
The sour fibre of tamarind pods is chewed by the children. Grown-ups soak the 
pods and make a sweet, cool drink from the liquid. 

Lemons are used in "achar" — pickle. The lemons are cut up, and ground spices 
are rubbed into the exposed inner surfaces. The spices used are cardamom, black 
pepper, fenugreek, and coriander. The lemons must remain covered with the 
spices for four or five days before being eaten. If prepared carefully, they will 
keep for a year. Plums are used to make a preserve. They are cooked, dried, and 
put into sugarcane juice. 

Mangoes are the utility fruit of the village. They are eaten raw, cooked with 
vegetables, and made into a variety of pickles and preserves. An oil pickle is made 
by cutting the mango almost in half and removing the stone. Aniseed, red pepper, 
and salt are ground and added to the fruit, and oil is added to coat the surface. A 
salty pickle is prepared by cutting the mangoes into four sections and removing 
the stones. They are put into a clay jar and sprinkled well with salt. Another 
form of mango preserve is made with sugarcane juice. Green mangoes are soaked 
and dried, or ripe mangoes are used, without cooking. They are dropped into a 
jar of strained, fermenting sugarcane juice, and left there for months or years. 
They are not ready for use until after remaining in the juice two months. Mangoes 
may be peeled, split, and dried after the stones have been removed. In this form, 
they may be added to almost anything which lacks flavour. A mango preserve is 
made by peeling the mangoes and rubbing them to a pulp with the spice-stone and 
roller. Salt, pepper, coriander and peppermint are added. This "chatni" (chutney) 
spoils quickly. 

Even the kernels of the stones removed from mangoes are used. The stones 
are split, the kernels taken out, boiled, dried and eaten. They may be roasted, if 

Animal products 

Milk and milk products are the chief forms of animal food in the village. Fresh 
milk is seldom used, unless there is an unusually large supply. When a cow first 
gives milk, the fresh milk is boiled, and clabbers almost immediately. This is 
eaten, with or without raw sugar. It is eaten during the next four or five days, as 
long as the milk is drunk after boiling. After this it is made into "ghi". In most 
homes all of the milk is reserved for "ghi". The milk is heated slowly in a clay 
jar, over a nest of smouldering dung-cakes. A little buttermilk is added, and it 
is set aside all night. By morning it has clabbered. At this stage it is called "dahi", 
and may be eaten with sugar. In city homes it is often served. But in the village, 
it seldom stops here, but goes on to butter. 



The "dahi" is put into the churn, water is added, and it is churned. "When 
butter forms, it is taken out and put into another jar. The butter is never used, 
but is collected for eight days. The sour buttermilk which remains in the churn 
has a variety of uses. The men of the family like it with their "satua" — parched, 
ground cereal — in the morning. It may be used with vegetables in "reota", de- 
scribed among vegetable dishes. Or it may be cooked with pulses in the dishes 
which we treat as meat substitutes. If there is not time to use it in any of these 
combinations, the men drink the buttermilk. Sometimes the children have a 
drink. The women do not drink it. 

The accumulated butter is heated in a deep kettle over the smouldering dung- 
cakes, or in the kettle for deep-fat frying over the fire-place. It may be heated 
from one-half to four hours, depending on the fire. Foreign matter which collects 
on the surface is skimmed off. When the heated butter is removed from the fire 
it is strained through a cloth. The clear fat which goes through is "ghi". It is 
regarded as the ideal medium in which to fry foods. Frying is an important 
process in village cooking, as it must be used for all foods served to guests and for 
food which members of the family must eat away from home, even though they 
eat it in their own fields. In the absence of ovens, many things are fried in "ghi" 
which we would bake. "Ghi" is also used in the preparation of a number of sweet- 
meats. In prosperous homes, it is an important ingredient in the special dishes 
given a woman two months before and a month after child birth. 

The Brahmans of our village eat no meat. They do not take advantage of the 
opportunity offered them by the Sacred Law, of eating meat as sacrificial food. 
Manu provides that "He who eats meat, when he honours the gods and manes, 
commits no sin, whether he has bought it, or himself has killed (the animal) or 
has received it as a present from others." V. 31. 

It may be that with the giving up of priestly duties, for farming, they have 
given up priestly privileges as well. If they were to eat meat, without sacrificial 
intention, they would be guilty of sin. "There is no greater sinner than that 
(man) who, though not worshipping the gods and manes, seeks to increase (the 
bulk of) his own flesh by the flesh of other (beings)." V. 52. 

The attitude of our Brahmans and of the orthodox Hindus who try to follow 
them is expressed in the Law. "Meat can never be obtained without injury to 
living creatures, and injury to sentient beings is detrimental to (the attainment 

of) heavenly bliss; let him therefore shun (the use of) meat." V. 48 "Having 

well considered the (disgusting) origin of flesh and the cruelty of fettering and 
slaying corporeal beings, let him entirely abstain from eating meat." V. 49. 

It is this attitude which keeps meat out of sight in the village. The Hindus 
who eat it, do their buying in the Muhammadan quarters or in their own shadow 
stables, as quietly as the excitement of bartering allows. The goldsmiths show 
themselves unorthodox in many ways, one of which is meat-eating. They do not 
talk about it but every one knows that they eat flesh. One or two carpenter 
families eat meat, as well as several families below them in the caste scale. They 

[Vol. 42 


all limit themselves to goat meat. The cow is revered too highly to be thought of 
as a source of food. Muhammadans eat as much goat meat as they can afford. 
Living in a Hindu community they have much the same attitude as their Hindu 
neighbours toward beef-eating. 

Hindus and Muhammadans prepare meat in the same way. They cut it up 
into small pieces, sear it in fat, add water and spices and cook it. They eat it with 
rice or bread. Meat dishes were the only ones denied me by my little Hindu 
cooking-teacher. She warned me that she would never touch our experimental 
fire-place or utensils if I were to defile them with meat. She knew that we ate 
meat, but that was not her responsibility, as long as it was kept out of her field of 


Sweepers and Dhanuks, two castes of outcastes, are the only ones who will 
touch pork. After watching the scavenging activities of village swine one is not 
surprised that their flesh is repulsive to most villagers. Only those who are brought 
up with the consciousness that they are themselves unclean, could eat it. The men 
do the butchering and divide the pig into shares. The women cut it up still 
further, into small pieces which can be fried and spiced, and eaten with bread or 



Fish is not classed with meat. Many eat it who abhor meat. We were sur- 
prised at the number of castes represented on the banks of drying ponds, waiting 
to start their "basket-fishing." On the occasional fishing days in the spring, the 
whole village smells of frying fish. Only Brahmans and those others who pride 
themselves on keeping the letter of the Law, abstain. 

Eggs are prohibited almost as much as pork. The fact that eggs hold potential 
life makes it impossible for orthodox Hindus to eat them. In so doing they would 
be destroying life. A further objection to the eggs is the objection to the chickens 
that lay them. Chickens are kept by outcastes. No one feeds them and they wander 
about the lanes near the outcaste section as scavengers. They are scrawny and 
offensive. The outcaste families who own them eat the few eggs which they lay. 


Special dishes are reserved for days of religious festivals, and weddings. From 
the passing of one festival to the coming of the next housewives and children, and 
often the men, recount the special dishes they have had or are going to have. A 
meal of toasted millet bread and stewed pulse tastes better if one thinks while 
eating it, of sweetmeats fried in "ghi". The pleasure of anticipation is granted to 
all, as those with much share with their dependents on these occasions. 

"Puris" — bread of finely ground wheat flour, fried in "ghi" — appear on all 
feast days. The other articles vary. "Holi" is the great festival of the spring. 
With its coming there appear "gojha", c yasse", "chandia", "puri", "kachauri" and 
"halwa". For "gojha", wheat flour is ground fine, sifted through a coarse cloth, 
kneaded and shaped into cakes as for bread. The cakes are dried for several hours, 
then fried in deep fat. When fried, they are pounded into small pieces and dried 



again. When well dried, they are ground in the stone flour mill, and put through 
the thread-strung sieve. The sifted material is combined with black pepper and 
crude sugar and again sifted. Raisins and "chiraunji" (small nuts) and cut up 
fresh cocoanut are then mixed with it. This is the filling. The crust is made of 
fine wheat flour, kneaded and rolled out as for small "puris". Each cake is flattened 
out in the hand and filled with the mixture. Then one side of the cake is folded 
over and the edges pressed together, very much like small turnovers. As rapidly 
as the turnovers are prepared they are fried in deep fat. 

"Yassee" is made of rice. The rice is husked, soaked, and pounded fine. It is 
then beaten up with a little warm <e ghi" or oil, and raw sugar is added while 

"Chandia" is made of split "urd." The "urd" is soaked and its husks removed. 
Then it is rubbed to a thick paste, together with asafoetida and salt. This paste 
is spread on a wet cloth in a small circle (about four inches in diameter). A hole 
is cut in the centre, and it is lifted from the cloth, fried in deep fat and then added 
to buttermilk. "Puri" and "kachauri" have been described under cereal foods. 

"Halwa" is prepared by making "puris" and grinding them up as for "gojha". 
One pound of this is browned well in one pound of "ghi", and combined with one 
pound raw sugar. A little water may be added if necessary. It is like a stiff mush, 
patted out and cut into cubes. Raisins may be added. In town there are many 
variations of "halwa", such as adding grated carrot, nuts and cocoanut. But in 
the village, the simple form is adhered to. 

In the spring, just before the last field of the winter crops is cut, the farmer's 
wife makes "sira". For this she browns wheat flour in "ghi". She melts raw 
sugar in a small quantity of water and adds hot water to it. Then she adds the 
browned flour and raisins and cocoanut and serves it. 

The wedding season comes in May and June. Wedding parties go from our 
village to be entertained in other villages and wedding parties from other villages 
are feasted by us. There are "puris", "kachauris," "halwa", and fried vegetables 
and pulse dishes, all of which have been described. A special dish is called "tikian". 
It is made of gram flour, mixed with enough water to make a stiff dough. In a 
deep kettle, water is heated — about three quarts water to one pound flour. The stiff 
dough is made into little flat round cakes, a little larger than the palm of the hand. 
These cakes are dropped into the boiling water and boiled about twenty minutes, 
until tough. They are then lifted out and spread on a tray to cool. Each cake is 
cut in half, and then into cross-wise strips. The gram flour inside looks packed 
and uncooked. Salt, red pepper and turmeric are ground and mixed through the 
strips. The sticky surfaces hold the spices. Two ounces of "ghi" and a pinch of 
asafoetida are heated in the deep-fat frying kettle and the spiced strips are fried 
in it. When the strips are well browned, water is added almost to the top of the 
kettle. It is cooked until most of the water boils away, when salt, cloves, carda- 
mom, and black pepper are added. It looks like noodles in thick brown gravy and 
tastes better than it looks. 

[Vol. 42 


Two festival dishes go through the same steps, but one is sweetened and one 
is made salty. The sweet cakes are "gul-gula" and the salty ones are "pakauri". 
To a pound of wheat flour a pound of melted raw sugar is added, also a small 
ginger root and a large cardamom. Water is added until the batter is like that 
for ginger bread. The frying kettle is half filled with "ghi" or oil, and heated. A 
handful of batter is dropped into the fat. The cakes thus dropped puff out and 
turn a rich brown. They are delicious, like doughnuts, with a strong molasses 
flavour. For the salty cakes, salt and asaf oetida are added to the flour and the same 
procedure is followed as above. We prefer the salty cakes. 

At wedding time mangoes are at their height, and appear in many forms. In 
addition to the preserves described, a special mango dish is prepared, called "Amiyan 
chinghora". Turmeric, pepper, salt and coriander are ground together. Oil is 
put into the frying kettle and the spices are heated in it. Gram flour and mangoes 
have previously been peeled and stoned. When it is ready to remove from the fire, 
heated asafoetida is stirred into it. 

In August, during the rains, there is a festival which is a happy one for the 
girls. Most brides, and often women long-married, go to their father's homes for 
this festival, "Raksha Bandan". The dish most definitely associated with this 
festival is "samai". Every woman makes it. She makes a stiff dough of the finest 
wheat flour she can grind. She inverts a large clay jar, and on its bottom she 
rolls the dough into a long thin string like vermicelli. The dried branch of a tree 
is stood on a roped cot, and the vermicelli is draped over it. When it is dry, the 
whole family enjoys it, mixing it with melted raw sugar. 

In the Autumn comes the Feast of Lights which is another occasion worthy of 
feasts. Puffed rice, plain or sweetened is the special dish for this festival. And 
there is "lapasi". A thin sweetened flour batter is poured into hot fat and stirred 
until it is a smooth thick brown sauce. This is poured out on a tray where it 
becomes firm almost immediately. It is patted out and cut into squares for serving. 
"Lapasi" may be made of wheat, maize, or barley, and may be sweetened or salted. 
Another dish associated with the Feast of Lights is "pitaua". Spiked millet or 
great millet is ground, and combined with raw sugar and sesame seeds. It is worked 
into a dough, and shaped piece by piece into round flat cakes, about six inches in 
diameter and three-eighths inches thick. These are fried. Maize may be used in 
place of the millet, in which case salt is added instead of sesame. "Halwa", 
described under the Spring Festival, also appears on the Feast of Lights. 

At any time during the year which is proclaimed auspicious, a man may give 
what is called "Katha". He calls in a priest to read from the sacred books, and 
invites all relatives within reach to come and listen, and feast. By giving a 
"Katha", a man gains much merit. At the feast he serves the guests mango pre- 
serve, with "Channa mirt". "Channa mirt J> is a mixture of cow's milk, leaves of 
the holy basil tree, Ganges water, sugar, curds, and gram. It is served in little 
clay cups, or poured into the hands of guests. At the same time, two wheat flour 
dishes are served. One of these is simply wheat flour mixed with hot "ghi" and 



then with raw sugar. For the other, wheat flour is made into a stiff dough with 
water and melted raw sugar. The dough is shaped into a long slender roll from 
which small slices are cut and fried in deep fat. 

Among the special dishes of the village must be included those prepared for 
women before and after their babies are born. One such dish is "mewa kilaurna", 
given to a woman during the early nursing period. Cocoanut, almonds, pistachies, 
raisins, dried dates, are all chopped together and heated in "ghi". Gum from the 
gum acacia tree is added. Raw sugar and water are boiled together until it forms 
a thread. It is stirred into the prepared condiments and the whole is beaten and 
shaped into balls. It resembles Persian sweets. This is only available in the better 


Another special food given a woman after child-birth is "harira". Aniseed, 
coriander, turmeric, fennel and "a j wain" (caryota copticum) are ground together. 
Equal weights of these and "ghi" are combined and added to their total weight of 
raw sugar. A little water is added. Both of these sweets are supposed to add 
strength, and increase the flow of milk. 


Certain practices in regard to food have become customary in Karimpur. They 
are the product of the experiences of generations. Some of them are common to 
all Hindu communities, others are limited to the Ganges area and still others are 
found only in our district. There are two sets of these practices. One set is asso- 
ciated with the good, and the other with the harm which may come to a person 
through food. Both have the support of custom, but those associated with the 
good which food may do, are such a simple part of daily living that one is hardly 
aware of them. They encourage the distribution and use of foods shown by ex- 
perience to be beneficial. Those practices associated with the harm which may 
come through food receive much more emphasis. They take the form of food 
taboos, supported by religious law and ceremony. 


Blunt, in his discussion of the caste system of Northern India, lists seven kinds 

of food taboos. (10). 

"The food taboos of Hindu life complicate it to an almost incredible 
degree: and as has many a time been pointed out, prevent c the growth of the 
good fellowship which we are wont to cement at the dinner table.* 
"Hindu food taboos are of several kinds: 


»o — w 


2. The cooking taboo— which lays down the persons who may cook the 


(10) See References. 

[Vol. 42 


3. The food taboo — which lays down what kinds of food a man may eat. 

4. The eating taboo — which lays down proper ritual at a meal. 

5. The drinking taboo — which lays down the persons from whom a man 
may take water. 

6. The smoking taboo — which lays down the persons whose pipe a man 
may smoke, and in whose company he may smoke. 

7. The vessels taboo — which lays down the nature of the vessels that a 
man may use for eating, drinking, and cooking." 

To a new-comer, these and related taboos seem an absurd waste of time. But 
as we observed the measures which we took to protect ourselves from disease in 
community with no sewers, no screens and no quarantine rules, we admired the 
skill with which high caste Hindus had formulated rules with all the prestige of 
ceremonialism, which served to protect them and their progeny from the dangers 
that apparently lurk in food. We tried to explain our reluctance to eat in village 
homes in terms of flies, and small-pox and cholera germs. But our village friends 
pleasantly ignored our terminology and explained our practices to each other in 
their own familiar terms of caste rules. They were relieved to find that although 
we mingled recklessly with all castes, we still maintained certain prohibitions. This 
discovery made us more comprehensible to them. 

The Brahmans at the top of the caste scale, are meticulous in their observances 
of the religious law. This includes bodily cleanliness. Even our busy Brahman 
farmers try to bathe every day. A man stands beside a well and pours water, 
freshly drawn, over his body, while he repeats prayers. He wears his loin cloth 
while bathing and changes it for a dry one if he has it, after the ceremony is com- 
pleted. After this bath he is ready to eat. No one is supposed to touch him until 
he has finished his food. Bose, in describing Hindu customs in Bengal, writes, 
"When Hindus sit together to eat, their seats are generally wooden planks or e Kusa' 
grass seats placed on the floor. These seats must not touch one another and no 
Hindu will touch another person who is also eating by his side. Among Brahmins 
this rule is very strictly observed. No one except the mother and the wife can 
touch a Brahmin while he is eating; if he does, he will stop eating and get up and 
wash his hands and mouth. This is a most barbarous custom. Some Brahmins do 
not allow their wives to touch them while they are eating. As the mother is re- 
garded as a goddess by all Hindus they make an exception always in her case." (11) 

The Brahman wife, who serves her husband, has also bathed or at least washed 
her hands before preparing his food. 

S. N. Jafri, Deputy Director of Public Information of the Government of 
India, gives an interesting picture of eating in a Hindu home of the United 

(11) See References. 



"The system of taking food by the Hindu tenantry, or for the matter of that all Hindus 
except those who are influenced by modernism, is unique. 

"They sit in a part of the floor of the kitchen which is called 'chauka* and is meant as a 
place for taking food. Before eating they take off all their clothes except the 'dhotees,' or the 
piece of cloth that goes round the waist and answers the place of trousers. In fact, other pieces 
are very seldom put on after bathing. The coat, the cap, and the turban, are all taken off, and 
a man eats with his body and head uncovered. Shoes are of course left at the door of the house. 
To woollen stuffs, they believe, there attaches no ceremonial uncleanliness and these they use 
while eating and worshipping, though they get them washed by bleachers, members of an inferior 
caste. In very cold weather, while eating, they generally throw a blanket over them. 

"The plates containing the food are put on the 'chauka,' because placing the food anywhere 
else would pollute it and then it could not be eaten. All the food that is cooked is not placed 
in the 'chauka' at once, but only the requisite quantity. The woman of the family who is the 
principal cook at the time, or officiates in the kitchen, serves the food for the eater or eaters. 
She sits near the fireplace, where the 'chauka' is situated. The man squats, in which case he has 
a small smooth board under him; his knees stick up close to his chest, the joint of the left arm 
and hand rest upon the left knee and it is thus the hand is supported while stretched out. 

"Both Hindus and Muslims eat with their fingers; knives and forks are unknown — the food 
is of such a kind that the fingers manage to carry it to the mouth, and thus they can eat very 
conveniently with one hand. 

"A Hindu at the time of eating must not be touched by one of inferior caste or by a non- 
Hindu for, if he were, he would immediately rise and not take another mouthful, even if he had 
to go without food the whole day; he would also throw out that which he might have in his 
mouth. He would never eat food prepared by a non-Hindu or even a Hindu of an inferior 
caste; and to some the shadow of a low caste man falling on their victuals renders them unfit to 
eat. A Hindu's food is of two kinds, 'kachcha khana' and 'pakka khana.' All the above-men- 
tioned ceremonies are required for 'kachcha khana.' No ceremonial uncleanliness attaches to 
dry things, such as flour and grain, and none also to fruits. Such things a man of higher caste 
can receive from a man of lower degree." (12) 

Before one can attempt to understand the intricacies of the food "Thou shalt 
nots" of a Hindu community, he must know the distinction between "kachcha" 
and "pakka" food. Ghurye in his presentation of Caste and Race in India, gives 
one of the clearest statements of this distinction that I have found. 

"Restrictions on feeding and social intercourse. There are minute rules as to what sort of 
food or drink can be accepted by a person and from what castes. But there is very great 
diversity in this matter. The practices in the matter of food and social intercourse divide India 
into two broad belts. Hindustan proper (our section of India), castes can be divided into five 
group; first, the twice-born castes; second, those castes at whose hands the twice-born can take 
'Pakka* food; third, those castes at whose hands the twice-born cannot accept any kind of food 
but may take water; fourth, castes that are not untouchable yet are such that water from them 
cannot be used by the twice-born; last come all those castes whose touch defiles not only the 
twice-born, but any orthodox Hindu. All food is divided into two classes. 'Kachcha' and 
Takka,' the former being any food in the cooking of which water has been used, and the latter 
all food cooked in 'ghi* without the addition of water. As a rule a man will never eat 'Kachcha' 
food unless it is prepared by a fellow casteman, which in actual practice means a member of his 
own endogamous group, whether it be caste or sub-caste, or else by his Brahmin 'Guru* or 
spiritual guide. But in practice most castes seem to take no objection to 'Kachcha' food from a 
Brahmin. A Brahmin can accept 'Kachcha' food at the hands of no other caste; nay, some of 
them, like the Kanaujia Brahmins, are so punctilious about these restrictions that, as a proverb 
has it three Kanaujias require no less than thirteen hearths. As for the Takka' food, it may be 
taken by a Brahmin at the hands of some of the castes only. A man of higher caste cannot 
accept 'Kachcha* food from one of the lower, though the latter may regale himself with similar 
food offered by a member of one of the castes accepted to be higher than his own." (13). 

(12), (13) See References. 

[Vol. 42 


Blunt gives the explanation of the distinction between the two kinds of food. 

"The Hindu draws a distinction between 'Kachcha' food, which is cooked in water, and 
Takka' food, which is cooked with *ghi' (clarified butter). This distinction depends on the 
principle that c ghi* like all the products of the sacred cow, protects from impurity: and since such 
protection is the object of all food taboos, this convenient fiction enables the Hindu to be less 
particular in the case of 'pakka* than of 'kachcha' food, and to relax his restrictions accordingly." 


In our village "kachcha" food is an exclusive, family-only product. But on 
certain definitely fixed occasions a Hindu may eat "pakka" food, prepared by others 
than members of his own family. The persons who may cook this food are all 
listed in the rules of his particular caste. In this way, although he ventures to 
dine abroad he is still somewhat protected. The persons who may cook the feast 
of "pakka" food, which he shares must be his equals or superiors in the caste scale. 
This guarantees that they will be at least as clean as he is, if not cleaner, in the 
care of their bodies and their food. If he is a Brahman, he is supposed to feast only 
at the house of a Brahman, with the understanding that the food has been prepared 
by members of the Brahman household, Brahman servants, or servants of water- 
carrier caste. He takes for granted that any sweets too elaborate to be prepared 


If doubt arises 

about the source of any article of food, the Brahman guest does not hesitate to 
make pointed inquiries of his host about it, before touching it. 

The following is just a short section from pages and pages of laws stating from 
whom a Brahman shall not accept food: 


The food of a king impairs his a Brahman's vigour, the food of a Sudra his excellence in 
sacred learning, the food of a goldsmith his longevity, that of a leather-cutter his fame. The 
food of an artisan destroys his offspring, that of a washerman his (bodily) strength." Code of 
Manu. IV. 218 and 219. 

The Brahmans of Karimpur treat their goldsmith neighbours as being worthy 
of becoming their hosts on special occasions. At the lower end of the caste scale 
are the Sudras. They can eat "pakka" food or drink water from members of the 
long line of castes above them. But better than any other food for the Sudra is 
that which a Brahman has touched. 

"Sudras who live according to the law, shall each month shave (their heads) ; their mode of 
purification (shall be) the same as that of Vaisyas, and their food the fragments of an Aryan's 
meal." Code of Manu. V. 140. 

Below the Sudras are the outcastes. They haunt the weddings and other feasts 
of any caste, ready to hurry in ahead of the dogs and crows, to gather up the 
scraps left by the guests on their leaf plates. These left-overs they 
baskets or ends of scarves or loin cloths. They eat what they can a 
dry out the rest to be eaten later. No one worries about protecting the outcastes 

from possible contamination 


(14) See References. 




Our years in India convinced us that experience had taught the law-makers 
well — at least for their own benefit — in the matter of "kachcha" and "pakka" 
food. We found ourselves unconsciously following their practice. We learned 
that we could safely go to a wedding feast in a village home, where we would hesi- 
tate to eat the daily food. This was because the wedding foods were all "pakka" 
fried in hot "ghi" and served at once. It was not the sacred source of the "ghi" 
which influenced us, but the fact that it was sizzling hot. 

The weighty distinction between "kachcha" and "pakka" food complicates 
food preparation for the village housewife. When her husband and sons eat in the 
family courtyard, everything may be "kachcha." She makes the ordinary cereal 
cakes and toasts them, and she may cook "sag" (boiled green vegetables) or pulses. 
If one of the men wants his food sent to the fields where he is working, she must 
prepare "pakka" food for him. She cannot send boiled food to the fields, as it 
would be "kachcha," and "kachcha" food may not be eaten in the fields where there 
is a possibility that the food or the eater might be touched. So she must fry the 
vegetables, and she is supposed to fry the cereal cakes in deep fat. These fried 
cakes are made only of pure wheat flour, which she cannot afford to use often. 
Neither can she afford large amounts of "gin" or oil for deep fat frying. So she 
toasts cakes of mixed flour in the usual way, and sprinkles a little "ghi" or oil over 
those which are to be carried to the worker in the fields, and pronounces them 
pakka". Thus she keeps the law with a few drops of "ghi". But her problem 
does not end here. If she should prepare her "kachcha" food such as toasted cakes 
and pulse first, then any food which she prepares after it becomes "kachcha", even 
though it would seem to be "pakka," by being fried in "ghi". It can then be 
eaten only by members of the family within the courtyard. If she has "pakka 
food to prepare she must finish it first, and set it aside in a prescribed place, and 
after that, cook the "kachcha" food. She finds it easier whenever possible, to 
have everything fried in "ghi" — "pakka", or everything "kachcha". 

"Kachcha" food must be eaten immediately. If a woman discovers that she 
has prepared more of some "kachcha" food, such as boiled vegetables or pulse, than 
will be used at noon, she can mud plaster a corner near her fire and set some of 
the food there to be kept for the evening meal of the same day. But she must do 
this before any one has been served. If any "kachcha" food is left after a meal 
has been served, the women and children may eat it later on in the day or even 
the next morning. The men will not touch it. One appreciates this precaution 
after observing what might walk over food standing on the floor, even though 
it be in a special corner, where no human treads. 

Drinking water is protected from contamination by a number of laws. Only 
three Brahman courtyards in Karimpur are large enough for wells. The other 
wells — 19 in all — are along the lanes and paths. Each well may be used by certain 
families, and must not be touched by others. Brahman women and other women 
kept in seclusion cannot leave their courtyards. Women of water-carrier caste 
draw their water for them. Whoever uses the well lets his or her jar down into 


[Vol. 42 


the water. This makes it important that unclean castes and strangers be prohibited 
from using a high caste well. We were not allowed to get our water from any of 
the village wells, as the boy who carried our water was a leather-worker, and the 
well of the leather-workers was too far away to be used. So our water came from 
an irrigation well in our grove and sometimes from one in the fields. We were not 
as particular as our village friends about who shared our well. In place of caste 
prohibitions, we resorted to thorough boiling. When cholera broke out in the 
village one summer, the district sanitary department sent out men armed with 
potassium permanganate. They did not limit themselves to the wells used by 
contaminated households but treated all wells, for safety. Someone told our 
people that the permanganate was poison. As we went the rounds to re-assure 
them we found men laboriously trying to empty the wells of the red water, a jarful 
of water at a time. If the permanganate had been somehow related to ceremonial 
purification or if waterboiling could have been substituted and given some spirit- 
appeasing significance, it would have been accepted readily. But being introduced 
officiously, without explanation, it was upsetting. 

Dishes in which food is prepared or served are carefully guarded. They are 
used for the immediate family only. Red clay dishes and clay cups without handles 
are used for guests and thrown away. They are never used a second time. This 
is as satisfactory as our rinsing with boiling water and it is much easier. At least 
so our young sons thought when we washed dishes after our own guests had de- 
parted. If clay dishes are too expensive, a host may serve his guests on plates of 
large leaves bound together with twigs. 

At a feast, nothing is ever passed around. The host or his sons serve the men, 
and his womenfolk serve the women, after the men have finished. As a server 
moves from guest to guest he is careful not to touch anyone with his garments 
and he is still more careful not to touch his hands, or the big iron serving-spoon 
against the plate of a guest. This would render the spoon, himself, and the serving 
dish in his hand unclean. 

Food dropped on the floor must not be eaten. The advantages of this are 
obvious, especially where the floors are the earth, and children and animals relieve 
themselves anywhere. The villager has his own explanation, expressed in an old 
saying, "If edibles at the time of eating drop down they are taken by the shadowy 
spirits and should not be picked up and eaten. If you do eat, you are sure to be 
possessed by the spirits who will harass you much." (15) 

There are other food regulations, the burden of which rests on the women. 
These have to do with the ceremonies of food preparation. In the morning as soon 
as her grinding is finished, a woman plasters her little fireplace and the floor all 
around it with fresh clay. The plastered floor space extends about three feet out 

(15) See References. 



from the fireplace. The clay she uses is not ordinary clay. It is collected from 
^the edge of a particular pond, by the children of the family if she is too high in 
caste to get it herself. She mixes the clay with a little water and a bit of cow- 
dung in her mixing bowl, and plasters it on the surface of the fireplace and floor 
with a rag kept for the purpose. When the plastering has been done, the fireplace 
and the plastered area around it are "purified". No one must go near them. Even 
the children of the family know that they must not approach it. It was not until 
I had my own fireplace that I appreciated the importance of this. In the absence 
of a kitchen table, it is necessary to have some place where one can put eatables 
and utensils and know that they will not be walked over. My fireplace was under 
a tree in our grove, with children playing all around, and farmers stopping to add 
their advice to that of my instructor. And only once did anyone approach this 
sanctum. That one was a little toddler of washerman caste. My teacher rushed at 
the child in a fury, and a torrent of reminders of her unworthy birth. The child 
fled, and we had no lesson that day. The fireplace had to go through a fresh mud- 
plastering before it could be used again. 

While the fireplace is drying, a woman may perform any of the tasks which 
do not demand strict purity. She may churn or make her fuel cakes of dung or 
chop vegetables. When everything is collected for the meal — flour, water, pulse 
or vegetable, and utensils and spices — she bathes. This she does by standing in a 
sunny part of the courtyard and pouring water over herself and her clothes. Then 
she changes to a "dhoti", a scant garment made of one piece of cloth draped around 
the body and over the head. She never allows the village washerman to touch this 
garment, but rinses it out at home herself. The washerman's touch would in- 
directly defile her, and the food she touches. Women lower than the goldsmiths 
in caste, let handwashing take the place of the bathing ceremony. 

When bathed, or washed, the woman enters the plastered area, by her fireplace, 
taking all of her materials and utensils with her. From this moment until after 
the men have been served, she must not leave it. The other women help with the 
preliminaries, but only the one who is ceremonially and physically clean, may 
attend to the final cooking. She alone may serve the men their food as they come 
and sit near her. In Karimpur the family never eats together, and even the men 
seldom eat together. Each one comes in when he is free, eats his meal in silence, 
and leaves the courtyard at once. The cook must wait until every man has been 
served before she is allowed to leave her post. Thus does a woman do her share in 
protecting the food which she gives her menfolk. For her own food, there are 
many less rules. She and the other women eat what the men leave. While they 
eat they are free to sit anywhere within the courtyard. Women are of little im- 
portance in the village interpretation of the Hindu code. 

[Vol. 42 



The second set of food practices, mentioned at the beginning of this chapter, 
are much less formal and less obvious than the ones just discussed. They have 
risen from experience which has demonstrated the good effects which apparently 
result from the eating of certain foods. Most of these practices are simply an ex- 
pression of physical desire for certain foods. I have called them "nutrition prac- 
tices," as they express the village way of meeting body needs. 

First there is the custom of sharing raw products. If there were barbed wire 
fences around orchards and fields our poorer families would suffer. As it is, each 
family of low caste with low food supply is dependant upon, and attached by 
custom to one or more high caste, prosperous families whom they call their patrons. 
When leguminous plants and mustard plants are young and their tops are tender, 
children of poor homes are allowed to pick them in the fields of their patrons. They 
gather enough so that their families have what they want to eat fresh, with some 
left over to dry and store for later use. The children also go into the fields of 
patrons and dig up certain of the green leaves which grow without tending, close 
to the ground. There are a number of those listed in the food table (p. 372). 

With a diet based on cereals and legumes, supplemented with vegetables which 
are chiefly of the pumpkin and tuber varieties, villagefolk need the minerals and 
the vitamins which these green leaves offer. And nature and custom have com- 
bined to make the leaves available, not just to the prosperous but to all. 

The same is true of sugarcane. From the time it is ripe until it is cut and 
pressed, one sees children and men of all castes sucking and chewing at the stalks. 
In so doing they are supplying their bodies with the calcium of which they get 
little in their low-milk diet. In the cane juice there are also the calories which they 

At harvest time sharing reaches its high point. Every man and boy in the 
village has a chance to work in the harvest fields. He may be a carpenter, a potter, 
or a leather-worker the rest of the year, but at this time he turns farm-hand. The 
untouchables are the only men excluded. We asked why? One of the farmers 
explained, "When they work just a little while, they perspire and then they com- 
plain because they are tired." We thought at first that this was an excuse for 
keeping the untouchables out of the fields. One expects to have real reasons veiled. 
But after seeing the untouchables at lighter field work, and their quick tiring, we 
agreed that the farmer had expressed his actual thought. A Multan proverb says, 
"He that eats a *ser' (from 2 to 3 pounds) works like a lion, but he that eats 
only a quarter of a 'ser' works like wood-ashes." And we wondered if this might 
not help explain the seeming laziness of the untouchables who have no fields of 
their own and who get only a meagre share from others. 

The men of all castes above the untouchables, and some of the low caste 
women, work daily while the harvest lasts. And every night each one carries 



home a headload of grain from the field just cut. This gives them varieties of 
grain and pulse which they would not otherwise have. 

At harvest time, women who cannot work all day, follow the harvesters, 
gleaning. A woman may glean in any of the fields belonging to her husband's 
patrons. In a few hours she can gather up several pounds of heads of grain- 
grain which would otherwise feed the crows. 

At harvest time representatives of families of craftsmen and serving castes 
appear at the fields of patrons at the proper time, to claim their shares. Even the 
outcastes have rights at this time. No questions are asked. Everyone knows whose 
duty it is to give, and whose right it is to receive, and the proper amount. The 
ethics of this system of payment may be questioned, but it assures the low caste 
and outcaste families of a variety of cereals and pulses which they would not 
otherwise get, either from their own fields or from the bazaar. 

In the fruit crops, we find a high degree of sharing. As has been noted, mangoes 
may be gathered by any member of the village, in any grove of the village. And 
mangoes are apparently their chief source of Vitamin C. The tamarind pods are 
available to all. And long ago Sanskrit medicine found in the tamarind, a quality 
now recognized as anti-scorbutic. 

The children are the ones who let their physical needs direct their foraging. 
None of the edible fruits miss them. The wild ones which they eat, we know 
little about. But we feel sure that they must supply some Vitamin C, and 
calories, and perhaps some of the iron which they need. The gum from the gum 
acacia tree of which they are particularly fond, must add calories, if nothing more. 
We tried most of the strange fruits which they brought us to sample. A few were 
delicious, but we would not recommend many of them for flavour. The only 
offering which we absolutely declined was something which they scraped from the 
trees of our grove. It was apparently the deposit of mud made by white ants to 
serve as their passageway up the side of the tree. What the children got from it 
which was satisfying, I cannot imagine. 

When the outcastes butcher a pig they treat the head as the choicest portion. 
It reminds us of the boar's head, praised by our own forefathers. The outcastes 
have not been told why the pig's head appeals to them especially. They only know 
that they like it. Here, as with the children and the fruits, nature guides them. 

The people of Karimpur have learned to save and eat many things which we, 
with our plenty, have wasted. It is comparatively recently that we have been 
taught that the water in which potaoes and certain other vegetables have been 
boiled is valuable. We try to save some of the water for soups. In Karimpur the 
water which is added to vegetables is so small in quantity that little or none re- 
mains when the vegetable is cooked. What does remain is eaten along with the 
vegetable. My village friends were horrified when they heard that in some places 
this water is thrown away. 

[Vol. 42 


They hate to throw anything away. When they cook vegetables, they never 
remove the skins, unless too tough for chewing. When their grain is threshed 
some of the finer husks remain. These come loose during the grinding of the grain 
into flour and are separated in the sifting. But the women are careful not to lose 
them. They put some back into the flour and save the rest to be mixed in with 
other flour which has less roughage. When I tell them that in America we sep- 
arate bran, and sell it in packages apart from the flour, at a price which seems 
fabulous to them, they think that we must be very stupid, or at least shockingly 

When a woman prepares a batter of any of the buttermilk or pulse mixtures 
in her mixing basin she does the stirring with her right hand, and scrapes the basin 
clean with the same hand. But she is not satisfied with this. She pours water into 
the mixing basin, washes the basin and her hand and pours the washed product 
into the frying pan or jar or other container along with the original preparation. As 
nothing is measured exactly, this additional liquid makes little difference. And 
the cook prefers to have it in the soup, rather than in her open mud drain which 
is already messy and alive with flies. 

As I walk through the village lanes on a festival day, I meet the women of 
serving castes, and outcaste women, going from door to door among their patrons. 
At each door they stand until one of the women of the patron's household tosses 
out several "puris" and the particular cakes or sweets associated with the occasion. 
If the serving woman thinks that she has not received her due, she says so, and 
stands talking about it until the women of the house give her something more, to 
get rid of her. If they think that she wants more than she should have, they 
shout at her until she is convinced that waiting will not profit her, and leaves. 
This sort of tumult alarmed me, until I learned that it was a perfectly friendly 
way of settling the amount of payment due. A definite payment in cash would 
seem more satisfactory to us. But cash, in the village is not as useful as food, 
especially the kind of food which is distributed on festival days. In the homes 
from which these low caste and outcaste women come, there is no wheat except 
that distributed at harvest time. They own no cows or buffaloes, and therefore 
have no "ghi". Meanwhile, there is plenty of wheat and "ghi" in the prosperous 
homes. And this method of handing them out in the form of festival cakes, as- 
sures the donors of religious merit, while giving the recipients and their children 
something which their bodies need. 

There are sometimes other rare things used, to fill the cakes, such as cocoanut 
and dried dates. These help still further to meet neglected needs. It seems like 
an unpleasant welfare method to us. But both parties regard it as an exciting and 
profitable procedure. Life would be dull without the anticipation of festivals. 
And as yet the West has not discovered a method of caring for the unfed, success- 
ful enough to be offered as an improvement. 



Among villagefolk there is a conviction that certain foods have special qualities. 
Some foods are "heating", while others are "cooling." We noticed that patients 
almost always asked what food they should eat. Knowing little about their foods, 
we hesitated to advise. But there was always someone around ready to speak with 
authority. If the patient suffered from toothache or earache, or a cold, the term 
"heating food" almost always occurred somewhere in the lengthy advice. If the 
patient had fever, "cooling food" always occurred. The reasons also seemed 
important. In cold weather when a person had digestive difficulties, "heating 
foods" were advised. In similar cases in hot weather, "cooling foods" were recom- 
mended. We collected a list of the foods which we heard mentioned in these two 
classes. "Cooling foods" include: buttermilk; sugarcane juice; bread of barley 
flour; tips of gram plants; "raskhir"; a combination of rice and sugarcane juice; 
radishes; carrots; Egyptian arum, or dasheen; and water-chestnuts. "Heating 
foods" include: bread of spiked millet; "dal" of pigeon pea; raw sugar; milk; 
potatoes; corn and sesamum oil. As I recall some of these foods, in the setting of 
an Indian hot season, I have reactions towards them, similar to those suggested by 
the villagers. But I have no adequate explanations of them as two distinct classes. 
There is no one quality running through either class which might explain its desig- 
nation. Further analysis of Indian foods may disclose a reason. At present, I 
simply offer the lists as demonstrating the effort of village folk to express the 
possible effects of food. 

Another class of foods accepted in the village is that of "strengthening" foods. 
These are supposed to give superior strength to the already strong, or to restore 
strength to the weak. The goldsmith brothers who act as village wrestlers, and 
are the acknowledged "strong men", eat almonds and pistachies, both of which 
are luxuries from the Mainpuri bazaar. While in training, they give up meat, and 
add milk to their usual diet. 

Strength- giving foods are considered important for expectant and nursing 
mothers. Two dishes reserved for them, "mewa kilaurna" and "Harira", are 
described in the section on festival dishes. These are too expensive for most homes, 
where simpler festival dishes of wheat or gram flour, rich in "ghi" and raw sugar, 
are substituted. Cocoanut and "phaphola" (nymphaea lotus) are included if 
possible, even though they must be begged. 

For anyone who has become very weak, without particular reason, mixtures 
like the two following are recommended by the local practitioners. Three drams of 
roots of the okra, and twelve black pepper corns, ground together. This is one 
dose. Or, one green seed pod of the gum acacia tree and ten black pepper corns. 
Grind and make into sherbet, with one cup of water. Sweeten to taste. Drink 
this every morning until strength returns. 

Sunshine is an acknowledged blessing in the village. Every morning good 
Brahman farmers chant hymns in its praise. We agreed with them, that but for 
the sunshine we would all have died. On several occasions, we found boys with 

crni11r^ v o^U„ „«-:i1 l:«~ ~:A\~r* nrl^ ^t-k^rc r\n r\\ir Wc W-J orm CiV tricvclc We 

[Vol. 42 


washed the playthings with disinfectant immediately. But we had to trust to the 
sun to save us from the germs which had had time to attach themselves before 
the discovery was made. 

In the rainy season one appreciates the sun, by its absence. After a few days 
of heavy clouds and rain, the village lanes are like pools, fed by the streams which 
trickle from open house- drains and privies. The smells which rise up to greet 
one are nauseating. A few days of bright sunshine clear the pathway and at- 
mosphere. The streams still trickle but they dry up before they become objection- 

In this discussion, however, we are more interested in the sun's light as a source, 
or producer of, Vitamin D. As far as we know, it is the villager's sole source of 
Vitamin D, except for a small amount which he may get from green leaves. I 
have heard Westerners inquire, "If sunshine is beneficial, then why aren't Indian 
children large and stronger than they are?" This is demanding too much of sun- 
shine. It cannot be expected to correct all diet defects. I have heard American 
doctors reply, that in rural India sunlight does well the work which they look to 
it to do. If our village children depended upon diet alone for adequate supplies of 
calcium and phosphorus, and for proper mobilization of the supplies, cases of 
rickets would be much more general and extreme than they are. I have found no 
definite records, beyond that of the work of Dr. Hutchinson and Dr. Shah. "A 
recent study of rickets in India by Hutchinson and Shah is clearly illustrative of 
the preventive role of sunlight. Surely there should be no rickets in India. Yet 
the infants of the top caste Hindus almost always develop extremely severe rickets. 
Owing to the religious custom of 'purdah,' which compels their mothers to live 
beyond the public gaze, these infants are practically never taken out of dusky 
rooms. In contrast the infants of Hindu laborers who cannot afford too expensive 
a religious detail practically never develop rickets. The women work with the men 
in the fields and leave their infants along-side in the sunlight. The rickets pre- 
ventive action of sunlight is here particularly clear cut, since in both instances 
the dietaries of the mothers almost entirely lack the food-stuffs which contain the 
substance D." (16). 

In Karimpur, custom co-operates with nature, in supplying sunshine to all 
castes. The women live and work in roofless courts. A few of these courtyard 
admit little sun in the winter months, but even they are better than stuffy rooms. 
Babies of all castes creep around naked during nine months of the year, and during 
the remaining three months they wear abbreviated cotton skirts. The children 
work in the fields or play in the lanes in scanty clothes which grow scantier with 
increasing heat. No one wears stockings, and few wear shoes. Craftsmen, instead 
of working in shops or work-rooms, carry on their crafts in village lanes or in 
open courtyards. The washerman, assisted by his whole family works beside a 
pond near our grove, not in a dark basement. Every one lives in the sunlight. 
And the sunshine blesses them. We have no cases nf spvprp richot* in the villaee. 

(16) See References. 



Some of the attitudes of our people toward food are expressed graphically in 
sayings and proverbs. These are an expression of their experience with food. Some 
of them are revealing enough to be included in this study of their practices. The 
rhythm of the lines is lost in translation. 



Juar (great millet) is my mother and makes my cheeks swell like raised 
sweet-cakes. Bajra (spiked millet) is my brother and restores my wasted form. 

"Whoever eats 'mung' pulse daily, becomes flatulent and drowsy." 

"Don't call me e arhar' (pigeon pea). My name is dear maiden. When all 
other grains are gone you come fumbling after me. Don't call me 'arharM My 
name is dear maiden. Two cakes of me are as filling as sixteen of other grains." 

"Arhar with dried mango and an ounce of 'ghi* makes a dish for the king." 

"Rice is good, but split peas are my life." 

"Everything is false in the world, save pulse and bread." 

"Eat bitter melons in September, radishes in August, and raw sugar in April 
this is the way to spend your money and buy illness." 

"Two things agree with a man — his own wife, and plain toasted cakes and 

"Two things bring pain— 'puris* (cakes fried in deep fat) and strange women." 

[Vol. 42 




The scientific study of food in India seems to have originated in the planning of 
diets for prisoners. In 1880, Surgeon-Major T. R. Lewis in his annual report 
points out a phase of the vegetarian diet of our people, which more recent students 
of nutrition have emphasized: 

"That chemical analysis, however exhaustive, can only afford such information as will enable 
a proximate estimation to be formed of the nutritive value of any food seeing that it is not only 
what nutriment a particular food-stuff contains that is of moment, but also what portion of it 
can readily be digested and assimilated by the body. In a diet composed entirely of vegetable 
substances, the quality of the cooking is of much more importance than it is in animal food 
dietaries, seeing that a large proportion of nutriment contained in cereals and pulses is enclosed 
in extremely resistant, indigestible envelopes, which if not effectually disposed of by proper 

cooking defeat all attempts on the part of the digestive organs to profit by the food 

In some cases, however, the excess of nitrogenous elements is given in the form of parched, or 
otherwise imperfectly cooked grain, so that it is probable that a large proportion of the con- 
tained nutriment will not be assimilated. On several grounds, therefore the addition of undue 
proportion of pulses — and especially ill-cooked pulses — is a doubtful advantage, and may be even 
injurious." (17). 

The most outstanding contributions in the study of jail dietaries were those 


His first investiga- 

tions were in Bengal, but it is his investigations into the jail dietaries of the United 
Provinces in which we are particularly interested. His statement of the jail 
dietaries in use in 1910 is useful in a consideration of diets of North India. 

"The diet scales for adult native prisoners on hard labour consist of 1 lb. 
pulse combination, 2 oz. of 'daP, 6 oz. of vegetables with salt, a little oil and 
pulse combination is invariably made up according to the subjoined table: 

12 oz 


of a cereal 
The cereal 












Spiked millet 
Great millet 


• • 

1 lb. 6 oz. 

• « 

1 lb. 8 oz. 

ft • 

1 lb. 8 oz. 

• • 

1 lb. 7 oz. 

• • 

1 lb. 7 oz. 

• • 

1 lb. 6 oz. 

• • 

1 lb. 6 oz. 

• • 

1 lb. 4J4 oz 











6 oz. 

4 oz. 

4 oz. 

5 oz. 

5 oz. 

6 oz. 
6 oz. 
7 J / 2 oz. 




will give spiked millet 12 oz. + maize 12 oz. -f pulse 4 oz. = 28, or 

will give wheat 11 oz. + barley 14 oz. + pulse 3 oz. 


"Any two of the above may be combined by taking half of each principal and half of each 
adjuvant; thus 1; 

2 . 

"These are the official diet scales as sanctioned in the Jail Code, but in practice they are 

to a large extent ignored. The only diet commonly in use is Diet E of the above table; in fact 

it is the only one given in the majority of the jails all over the Province. During the cold 

weather months great millet, spiked millet, and barley are sometimes made use of, but hardly 

ever in the quantities laid down. The different superintendents give their own combinations of 

these food-stuffs with wheat and pulse, which are usually very different from the diets officially 

(17) See References. 



"Rice has been very rarely used in the United Provinces. As it is only grown to a limited 
extent, the price is prohibitive. It was, however some years ago used in two district jails, close 
to the borders of Bengal, but an outbreak of Beri-Beri in both of these caused it to be dis- 
continued and its use has never been revived. 

M 'Mania' (millet) is never given at all; in fact, we failed to obtain samples even for analysis 
from the different jails, where investigations were carried out. 

"Maize is seldom, if ever used: why this is so we do not understand as it is a very nutritious 
food-material and certainly superior to some of those in use. 

"The pulses made use of in the United Provinces are principally 'arhar dal' (pigeon pea) 
and 'urd dal.' It would appear from the above table gram 'dal' is not classed as a pulse, but 
this is mistake as gram 'dal' belongs to the pulses. 

1 lb. 7 oz. 
5 oz. 

2 oz. 

3 oz. 

150 grains. 
9.08 grams. 
2 oz. 

"For at least nine months in the year Diet E, i.e 


Gram . . • • 

Arhar dal (pigeon pea) 
Vegetables . • 


Condiments . . 

is given to the exclusion of all others." (18) 

The diet now in use in our Mainpuri District Jail is practically what McCay 
recommended as a result of his investigatons. 

In jail, the prisoners have a cereal, a pulse, and a vegetable daily. Our village 
people are glad if they can have a pulse or a vegetable with their cereal bread. They 
rarely have both on the same day. And sometimes they have neither, and eat salt 
with their bread. But the villager spends his time in the fields where he may snip 
off the green tips of plants or roast a handful of grain. The prisoner's fare is 
exactly that which is recorded. The farmer's is very likely not exactly what we 

have recorded. 

McCay has some interesting comments on the contents of jail diets as compared 

with diets of the province outside the jail. 

"We may, therefore, conclude this consideration of the pulses of these diet scales by stating 
that they are present in quantities much in excess of what is required, and in far greater 
abundance than the people of the Province can afford, or wish to have. Two and a half chittacks 
or five ounces is the maximum amount of pulse that should enter into any diet, and, if gram 
is to be given it should be properly cooked, not made use of as parched gram, and it should 
never be the only pulse in the composition of the diet. The superiority of 'arhar dal (pigeon 
pea) and the fact that it is the favourite form of pulse with natives of the United Provinces 
should ensure that this pulse will be used in all diets to as great an extent as possible. Urd dal^ 
being more suitable than other forms in the preparation of 'bajra' (spiked millet) and ujar 
(great millet) for consumption, this Mai' should be used in diets into the composition of which 
those cereals enter. One chittack (2 ounces) of «urd dal' will be found amply sufficient in the 
cooking of the quantities of 'juar* or 'bajra' recommended in new dietaries." (19). 

"Diet C in which the principal is 'makka* or maize, we have not investigated. 
It is never used, so far as we could gather, in the dietaries of the prisoners. Maize 
we found in Bengal was a very good food-material although not very palatable as 
prepared in Bengal Jails. It is not suitable alone for baking into bread or 'chapatti,' 

very good 


think it suitable as a principal, as in Diet C, but see no reason why maize should 

(18), (19) Sec References. 

























•a © 












© r 3 T ? r c!#'d 


I S fl ® 

o3 £ g 00 S 


© N 




in „w O N 

e *£|* * 

•f J 1 1* -i 


1 b S S 




not be given as a substitute for part of the wheat in dietaries. Why it should 
never be given in the jails of the United Provinces is difficult to understand, as, 
next to wheat, maize is the best cereal on the list. It is used during its season 
fairly largely by the inhabitants of the province, and the average crop being up to 
one million tons, there cannot be any great scarcity of this food to enhance its 
price. We are strongly of the opinion that maize has been neglected unjustly, and 
that, if given as we suggest, mixed with wheat, it would be found quite satis- 
factory." (20) 

"Judging from inquiries we have made, wheat is the food of the higher and 
middle classes and a luxury to the poorer. In times of scarcity or famine, when 
wheat, gram and 'juar' are practically the same price, wheat is eaten largely by 
the people, but in ordinary times e bajra' and 'juar' are the principal food-materials 
of the poorer classes and small cultivator. 

"Makka or maize is also largely used as a food-material, barley and 'mania' not 
to so great an extent as the other cereals. c Arhar daF is the favourite form of 
pulse, and a small amount is partaken of daily. 

"It is, therefore, only right and fair to the prisoners that these secondary food- 
materials, 'bajra,' 'juar/ 'makka,' barley and c marua,' should not be denied them. 
While they are not suitable to be used alone, when given along with wheat, in the 
manner we have indicated in the diets suggested, they are exceedingly useful in 
varying the monotony of the wheat diet, and in providing food-materials to which 
the great mass of the prisoners are accustomed. (21). 

" from a physiological point of view it must be admitted that the dietaries 

at present in force, or any framed in future to maintain the same level are dis- 
tinctly superior to those available for the great mass of the population, and that 
this will tend to place a premium on crime in periods of scarcity. This, how- 
ever, is a problem with which we have no concern; we simply point out its bearing 
in passing and leave the present policy of the maintenance of the superiority of 
jail dietaries, over those possible to the same classes outside the walls of the jails, 
to be determined by the government of the country." (22) 

We hope that our people will find an adequate diet without being obliged to 

go to jail to find it. 

McCay's investigations led him to the conclusion that the protein factor 
determines the place of Indian races in the scale of physical efficiency. He traced 
all differences in physique to the different levels of nitrogenous interchange possible 
in the diet of wheat-eaters and the diet of rice-eaters. If he had made his investi- 
gations a few years later he would have modified his conclusions to include minerals 

and vitamins. 

Since the studies of McCay there has been an increasing interest in the study 
of foods in India as elsewhere. But very little scientific work has been done. 
Colonel R. McCarrison has contributed more than anyone else through his experi- 
ments and through his efforts to use the results of his experiments in educating 
Indians in the possibilities of indigenous foods. He was a member of the Indian 
Medical Service until 1919 when he became Director of the Deficiency Diseases 
Inquiry, in the Pasteur Institute, Coonoor. 

(20), (21), (22) See References. 

[Vol. 42 


One experiment which he carried on with two colonies of rats (20 in each 
colony) contrasted what he calls a "good diet" with a "bad diet." "One colony 
received a 'good diet/ designed to resemble that eaten by the Sikhs. It consisted 
of 'chapatties' made of whole wheat flour; uncooked, green vegetables, fresh fruits 
(tomatoes, etc.); sprouted gram (legumes); butter; fresh whole milk; and fresh 
meat occasionally." (23) Liberal quantities of each article were supplied so 
that the animals could select for themselves the amounts of each they cared to eat. 

This diet is of interest to us, because all of the foods in it are available in our 
part of the country. Our village folk grow wheat, but they sell much of it. They 
do not grow as many green vegetables as they might, but at the same time they 
make use of every edible leaf within reach. They have plenty of gram, but they 
are accustomed to parching or cooking it. It would be easy for them to sprout it. 
They could grow tomatoes and oranges and lemons. Prejudice keeps them from 
growing tomatoes, and they have not tried growing citrus fruits, other than a few 
lime trees. Oranges are now grown in our area of the Province, and can be grown 
in Karimpur. They would be in season during the months when there are no 
mangoes. The village folk might have butter, but they clarify it all to make 
"ghi." The same is true of milk. There is milk in almost every home and some 
families drink it. But in most households the milk is heated and churned to be 
made into "ghi." As for meat, a few eat it now. The meat at present sold is very 


tion for meat-eating. 


McCarrison's "poor diet" 


people of the poorer classes— white bread, cooked vegetables, a cocoanut oil mar- 
garine, canned meat and jam, with preservatives, and tea, with milk and sugar. 

The experiment lasted six months, during which time all of the rats were given 
equally good care. They all had a two-hour sun bath daily. During the six 

rats out of the 20 on the "good diet" died, one of them from injury- 
Nine of the 20 rats on the "poor diet" died. Three of these were killed and eaten 
by their fellows. After this, the rats of the ill-fed colon v were «rar*t*d 



night and a small quantity of fresh vegetables was given three times a week. The 
other six died from broncho-pneumonia. There were 20 litters of rats born in the 
well-fed colony, and all of the 134 born were reared to maturity. In the ill-fed 

Two of these 


eleven died and the others were eaten before they could be removed. 

The aggregate body weight in each colony was 2,540 grams at the start. At 

the aggregate weight of the "well-fed" colony (17 


was 1,300 grams. 

Colonel McCarrison 


docs the former (good diet) promote 

frs^r j\.«.\ • • . X6 ' F rOT1 <>ie Physical efficiency and health but the latter 

tinal disease." (24) 

(23), (24) Set References. 



This last statement is based on his post mortem examinations. 

Our people have the symptoms of the bad diet, while the good diet is within 
their reach. Experience has taught them much. But Science can add what ex- 
perience has evidently missed. Few of them can read and those few see no news- 
papers or journals. The contributions of Science have no way of reaching them 
as yet. With its help they will be enabled to build up stronger bodies, without the 
need of importing new foods. With the raw materials already available, they 
have a better selection of foods than people in the West who are not as poor as they. 

An article by Colonel McCarrison which appeared in "The Practitioner" on 
"The Relation of Diet to the Physical Efficiency of Indian Races" (25) is of value 
in our study. It is worth quoting at length. 

"Few who have travelled far in India can have failed to notice the remarkable differences in 
physical efficiency of different Indian races. So great is the contrast between certain stalwart, 
vigorous, and resolute people of the North and certain poorly developed and toneless inhabitants 
of the South and East, that the question arises: Why should there be these great differences 
between one race and another? In attempting to answer it, I shall put aside such factors as 
climate, prolonged exposure to the actinic rays of the sun, peculiar religious customs and endemic 
diseases. For while it cannot be doubted that these play their part in determining the position 
of different Indian races in the scale of physical efficiency diet is the most important of the 
factors concerned 

"The food materials available to the people of India ... are mostly derived fom the vegetable 
kingdom. They include rice, wheat, barley, maize, millet, legumes, vegetables and fruit. For 
various reasons — of which locality, poverty, religious scruples and habits are the chief — milk and 
milk products, eggs, and flesh meat enter little, if at all, into the diet of great masses of the 
people. The use of these commodities is confined, as a general rule, to the better classes and to 
certain races in Rajputana, the Punjab, Northern India, and the West Coast. The stapk food 
of the masses thus consists of 'those parts of plants which have the function of storage tissues' 
(McCollum), viz., cereals and legumes. 'These are deficient in at least three dietary factors. 
All contain protein of relatively poor quality; all contain too little of certain mineral elements, 
especially calcium, sodium, and chlorine and all are deficient in fat-soluable vitamin A. It is 
true that the diet is supplemented with vegetables and fruits, but these are not eaten in sufficient 
quantity to compensate for the defects of the cereals and legumes. For these reasons the diet 
of many millions of people in India is an incomplete one which 'does not contain all the com- 
plexes necessary for the construction of living protoplasm' (McCollum). Now it is, to a large 
extent, in proportion to the way in which the defects of what may be called 'the basal diet' of 
the people of India are made good by the use of other foods containing more suitable protein, 
more of the essential mineral elements, and more vitamins, that the physical development and 

well-being of the different races of India depend. 

"The least satisfactory of all the diets in common use in India is that consisting of rice, 'dal' 
(legumes), vegetables and condiments. This diet is used by millions of people in Bengal, Madras, 
and elsewhere; and it is these who exhibit the lowest grade of physical efficiency and health." 

We need not follow in detail, as it does not apply to our section of the country. 

"If now it be found that the diet of other races is such as corrects some or all of the five 
faults pertaining to the diet of rice and legumes, then, other things being equal, the better diet 
ought to be associated with a higher grade of physical efficiency in the races using it; this is 
precisely what we find. For when we pass from the rice and legume eaters of Lower Bengal to 
the wheat and legume eaters of the United Provinces— climate and other factors being practically 
the same in both localities— we see that the latter are 'better developed physically, more capable 
of hard work, hardier and more alive than the average resident of Bengal of the same class 
(McCay). The reasons for this are that wheat provides a protein of better quality than rice; 
the composition of the diet admits of the absorption of a larger amount of protein, of the attain- 
ment in short of a higher level of nitrogenous interchange (McCay) ; there is not the same 
ingestion of carbohydrates nor the same tendency to gastro-intestinal disturbance; the deficiency 

(25) See References. 


[Vol. 42 



of mineral elements is not so great, and the diet is, in general, richer in vitamin B. It is, how- 
ever, faulty in regard to the quality of its protein and to its content of mineral essentials and of 
vitamin A; and since its users do not, as a general rule, supplement it with a sufficiency of other 
foods, rich in those elements and complexes in which it is poor, they lag behind, in physical 
development and well-being those races that do." 

This describes the people of Karimpur